Skip to main content

Full text of "Industrial organic analysis, for the use of technical and analytical chemists and students;"

See other formats


INDUSTRIAL 
ORGANIC     ANALYSIS 

FOR  THE  USE  OF  TECHNICAL  AND  ANALYTICAL 
CHEMISTS  AND  STUDENTS 


BY 

PAUL    S.    ARUP,    B.Sc.,   A.C.G.I. 

WITH    A    FOREWORD 

BY 

J.   C.   IRVINE,   D.Sc.,   Ph.D. 

PROFESSOR    OF    CHEMISTRY    AT    THE    UNIVERSITY    OF    ST.     ANDREWS 


With  fourteen   Illustrations. 


~*u*. 


PHILADELPHIA 

P.    BLAKISTON'S    SON    &    CO. 

1012  WALNUT  STREET 


Printed  in  Great  Britain. 


FOREWORD 

BY  PROFESSOR  J.  C.  IRVINE,  D.Sc.,  PH.D. 

THE  relationship  between  science  and  industry  has 
been  the  subject  of  endless  discussion  in  recent  years, 
and,  if  tangible  results  have  not  been  numerous,  signs  are 
not  wanting  that  the  value  of  scientifically  trained  men 
in  industry  is  now  frankly  recognised  in  some  quarters  and 
that  the  universities  and  colleges  are  anxious  to  provide 
for  any  demand  for  such  men  which  may  arise. 

So  far  there  is  agreement  between  the  parties,  but, 
assuming  that  the  employer  is  perfectly  clear  as  to  what 
he  wants,  teaching  institutions  are  by  no  means  agreed 
on  the  best  means  of  modifying  their  curricula  to  meet 
modern  requirements.  The  orthodox  course  for  a  degree 
in  science  certainly  shows  a  desirable  uniformity,  but 
differences  of  opinion  at  once  arise  when  attempts  are 
made  to  frame  an  educational  policy  which  will  retain 
for  institutions  of  university  rank  the  ideal  of  wide 
culture,  while  admitting  some  aspect  of  practical 
training  of  use  in  the  field  of  industrial  work. 

In  the  particular  case  of  chemical  training,  these 
difficulties  are  modified,  if  not  entirely  removed,  when 
the  special  qualifications  desirable  in  the  technical 
chemist  are  taken  into  account.  That  he  should  have 
received  as  thorough  a  training  as  possible  in  what  is 
often  termed  "  Pure  Chemistry  "  may  be  accepted  as  a 
matter  of  course.  It  is  equally  essential  that  the  student 
should  be  trained  to  think  scientifically,  and  that  his 


iv  Foreword 

capacity  for  original  inquiry  should  be  developed.  Such 
a  result. may  be  achieved  in  particular  instances  through 
the  inspiration  of  a  gifted  teacher,  but  the  best  route  to 
this  desirable  end  is  found  in  a  course  of  original 
research  undertaken  after  graduation  or  its  equivalent. 

It  may  be  admitted  that  the  young  graduate  is 
frequently  of  little  immediate  value  to  the  manu- 
facturer. It  is  equally  true  that  he  has  to  be  trained, 
often  slowly  and  laboriously  in  the  methods  and  spirit 
of  research,  but  it  is  at  once  the  privilege  and  duty  of 
the  university  or  college  to  supply  this  training.  After 
two  (or  it  may  be  three)  years  spent  at  this  work  the 
latent  initiative  and  resource  of  the  student  are  brought 
into  operation,  and  his  capacities  for  useful  employment 
are  thereby  increased  enormously.  In  the  absence  of 
any  definite  experience  in  technical  methods,  however, 
these  powers  will  not  be  developed  to  their  fullest 
advantage.  Strictly  speaking,  a  course  of  specialised 
technical  work  should  follow  at  this  stage,  but  the 
obstacles  in  the  way  are  manifold.  Our  student  is  now 
probably  between  twenty-four  and  twenty-five  years  of 
age,  and  is  naturally  anxious  to  begin  his  life's  work. 
Unless  he  has  good  prospects  of  obtaining  a  position  in 
some  particular  industry,  he  is  hardly  justified  in  making 
a  choice  of  any  one  branch  of  Applied  Chemistry  for 
special  study.  Obviously  in  the  course  of  training  here 
outlined,  the  shortest  possible  route  to  technical 
experience  has  to  be  taken. 

In  most  teaching  institutions,  little  or  no  provision  is 
made  for  instruction  in  Applied  Chemistry,  or  in  the 
technique  of  Commercial  Analysis  as  a  course  to  be 
taken  only  after  scientific  Chemistry  has  been  absorbed 
by  study  and  expanded  by  research.  We  have  the 
university-trained  chemist  and  the  technical  chemist, 


Foreword  v 

the  former  with  a  slender  idea  of  the  world  of  manu- 
facture, and  the  latter  often  lacking  the  mental  discipline 
of  pure  research,  and  each  type  of  man  requires  some- 
thing more  before  he  can  be  really  useful. 

Mr.  Arup's  book  is  an  attempt  to  supply  the  deficiency 
of  one  of  the  above  types.  He  has  provided  a  course  of 
technical  organic  analysis  which  can  be  profitably  under- 
taken either  by  the  undergraduate  or  by  the  more  mature 
original  worker.  The  author  has  directed  his  attention 
to  modern  standard  methods  which  apply  to  important 
substances,  and  .the  best  type  of  student  who  works 
through  a  selection  of  the  exercises  should  get  into 
touch  with  technical  methods  and  the  literature  of  com- 
mercial processes.  Whatever  business  instincts  he  may 
have  will  be  aroused,  and  some  definite  form  and 
direction  conveyed  to  his  efforts  to  obtain  a  suitable 
post. 

A  comparatively  small  amount  of  preliminary  tech- 
nical training  can  often  transform  the  research  chemist 
into  a  valuable  official  in  a  business  concern  or  factory, 
but  some  beginning  must  be  found  for  this  transforma- 
tion. As  Professor  Martin  Bogert  has  recently  said,* 
"  the  really  important  thing  is  to  bring  together  the 
problem  and  the  man  competent  to  solve  it,"  and  although 
no  laboratory  course  can  take  the  place  of  factory  expe- 
rience, the  training  embodied  in  Mr.  Arup's  book  may 
help  many  young  chemists  to  offer  themselves  for  tech- 
nical posts  with  some  knowledge  of  how  to  begin  their 
work,  and  some  ideas  regarding  its  possible  develop- 
ments. 

Apart  from  its  effect  on  the  student,  there  is  one  other 
aspect  of  the  training  indicated  in  this  book,  and  that  is 

*  Presidential  Address  to  the  Society  of  Chemical  Industry, 


vi  Foreword 

its  influence  on  the  teacher  of  Chemistry.  In  the  address 
already  quoted,  Professor  Bogert  emphasises  this  point 
and  states  that  if  the  teacher  were  more  familiar  with 
industrial  problems  he  would  often  select  topics  for  his 
research  students  bearing  on  these  subjects.  The  adop- 
tion of  such  a  policy  would  in  no  way  endanger  the 
progress  of  scientific  Chemistry  in  face  of  the  fact  that 
the  large  majority  of  students  of  Chemistry,  who  must 
inevitably  qualify  themselves  for  technical  posts,  would 
in  the  natural  course  of  events  carry  out  some  strictly 
scientific  research  as  a  preliminary  to  technical 
investigation. 

Mr.    Arup's  book   is   a   practical    contribution    to    a 
problem  which  becomes  more  urgent  every  day. 

J.  C.  IRVINE. 
THE  UNIVERSITY, 

ST.  ANDREWS, 


PREFACE 

THIS  volume  has  been  written  for  the  use  of  students 
who,  having  received  a  thorough  training  in  theoretical 
and  practical  Chemistry,  may  desire  to  gain  some 
insight  into  the  methods  and  principles  of  industrial 
organic  analysis,  and  as  a  work  of  reference  for  chemists 
engaged  in  technical  or  analytical  work. 

Considering  that  the  great  majority  of  students  who 
specialise  in  Chemistry  sooner  or  later  take  up  indus- 
trial work,  the  Author  considered  that  some  attempt 
might  be  made  towards  bridging  the  gap  which 
undoubtedly  exists  between  academic  and  scientific 
Chemistry,  on  the  one  hand,  and  industrial  Chemistry 
on  the  other.  With  this  motive,  he  has  written  the 
present  work,  endeavouring  to  treat  the  subject  in  such 
a  way  that  it  may  prove  of  interest  from  the  scientific, 
industrial  and  educative  points  of  view. 

It  may  be  pointed  out  at  once  that  the  subject-matter 
of  this  volume  is  not  intended  to  replace  any  portion  of 
the  ordinary  curriculum  or  the  research  training  which 
forms  so  valuable  a  feature  in  the  educational  pro- 
grammes of  many  of  our  colleges  and  universities ;  it 
is  recognised  that  the  academic  training  of  the  industrial 
chemist  should  be  as  thorough  and  comprehensive  as 
possible,  in  order  that  he  may  be  able  to  take  a  broad 
view  of  the  many  different  problems  which  may  con- 
front him  in  practice  ;  it  was,  however,  thought  that  the 
present  system  of  training  might  well  be  supplemented 
by  showing  the  student  how  his  theoretical  knowledge 


viii  Preface 

can  be  applied  in  industrial  Chemistry,  at  any  rate,  as 
far  as  his  laboratory  work  is  concerned. 

There  is  no  doubt  that  the  student  just  entering  on 
industrial  work  has  much  to  learn,  and  that  his  success 
will  generally  depend  on  the  quickness  and  efficiency 
with  which  he  applies  his  knowledge  to  the  solving  of 
the  problems  with  which  he  has  to  deal ;  the  object 
aimed  at  here  is  to  help  to  give  the  student  some 
opportunity  of  putting  himself  into  touch  with  the 
requirements  of  an  important  branch  of  his  new  work, 
and  at  the  same  time,  it  is  hoped  that  something  may 
be  done  towards  modifying  the  somewhat  sceptical 
attitude  sometimes  taken  by  the  manufacturer  towards 
the  scientifically  trained  man  or  rather,  the  student 
fresh  from  college. 

Regarding  the  scope  of  this  work,  it  may  be  said  that 
only  such  sections  of  the  subject  as  seemed  to  be  of 
sufficient  general  interest  have  been  chosen.  More 
attention  has  been  paid  to  natural  products  and  sub- 
stances obtained  from  these  by  comparatively  simple 
means  than  to  products  which  are  obtained  by  complex 
manufacturing  operations.  The  object  has  not  been  to 
teach  certain  methods  of  analysis  connected  with 
specific  manufacturing  operations,  but  rather  to  initiate 
the  student  into  the  general  methods  and  spirit  of 
industrial  organic  analysis.  He  may  thus  learn  to  cope 
with  more  complex  problems  in  this  field  of  work  with 
the  aid  of  the  standard  works  of  reference. 

Regarding  the  choice  of  analytical  methods,  the 
Author  has  selected  standard  well-tried  processes, 
especially  those  which  are  instructive  in  throwing  light 
on  the  characteristic  chemical  behaviour  of  the  sub- 
stances under  examination.  Methods  requiring  the  use 
of  special  apparatus  such  as,  for  example,  the  refracto- 


Preface  ix 

meter  or  coal  calorimeter,  not  usually  to  be  found  in 
the  ordinary  teaching  laboratory,  have  not  been  de- 
scribed in  detail ;  through  the  references  given  both  in 
the  text  and  at  the  end  of  each  chapter,  however, 
the  reader  will  readily  find  any  such  methods  fully 
described. 

More  complete  working  details  have  sometimes  been 
given  than  would  be  required  by  the  expert  analyst, 
though,  on  the  other  hand,  it  has  not  been  thought 
necessary  to  describe  in  detail  such  well-known  ana- 
lytical operations  as  may  be  found  fully  described  in 
most  text  books  on  organic  and  inorganic  analysis. 

Stress  has  been  laid  on  the  reasons  for  which  every 
analytical  operation  is  carried  out,  as  well  as  on  the 
discussion  of  the  practical  significance  of  the  results 
obtained.  In  this  way,  it  is  hoped  that  the  work  may 
prove  interesting  to  students  and,  at  the  same  time, 
impress  them  with  some  of  the  most  important  aspects 
of  industrial  analysis. 

Frequent  references  have  been  made  to  standard 
works  of  reference  when  the  subject  passes  beyond  the 
scope  of  the  work,  and  it  is  hoped  that  the  bibliography 
appended  to  each  chapter  may  prove  a  useful  addition 
to  the  text.  To  these  works  the  Author  desires  to 
express  his  indebtedness  for  much  of  the  material  con- 
tained in  this  volume,  and  he  also  wishes  to  thank 
Professor  J.  C.  Irvine,  of  the  University  of  St. 
Andrews,  for  valuable  suggestions  received  in  the 
course  of  its  preparation. 

The  Author's  thanks  are  further  due  to  Mr.  W.  A. 
Davis,  of  the  Rothamsted  Experimental  Station,  and 
Mr.  S.  H.  Blichfeldt,  chief  chemist  and  bacteriologist  to 
Messrs.  Otto  Monsted,  Ltd.,  for  useful  information 
concerning  analytical  methods,  and  to  Mr.  R.  E. 


x  Preface 

Doolittle,  of  the  New  York  Foods  and  Drugs  Section 
Laboratory,  for  his  courtesy  in  supplying  information 
concerning  the  law  in  America  regarding  preservatives 
in  foods. 

PAUL  S.  ARUP. 

SOUTHALL, 
1913 


The  Author  is  indebted  to  Messrs.  P.  Funke  6-  Co.,.  and 
to  Messrs.  A .  Gallenkamp  &  Co.,  for  the  loan  of  blocks  for 
some  of  the  illustrations  appearing  in  this  volume. 


CONTENTS 


FOREWORD iii 

PREFACE     .         .        .         .        .  ...         .         .      vii 

INTRODUCTION I 

CHAPTER   I 

COAL    AND    COKE 

Classification  of  Coals— Analysis  of  Coal  and  Coke         .        7 
CHAPTER   II 

COAL    TAR    AND    ITS    DISTILLATION    PRODUCTS 

Production,  Composition  and  Examination  of  Coal  Tar 

and  its  chief  Distillation  Products  33 

CHAPTER   III 

THE  FATTY  OILS  AND  FATS 

Composition  and  Classification  of  Oils  and  Fats — General 
methods  for  their  Examination- — Analysis  of  simple 
Fatty  Mixtures  ........  79 

CHAPTER   IV 


Production — Composition  and  Classification  of  Soaps — 
Examination  of  Soaps — Analysis  of  Phenolic  Dis- 
infectants containing  Soap  »  .  .  .  125 


xii  Contents 


CHAPTER  v 

PETROLEUM    AND    ITS    DISTILLATION    PRODUCTS  — 
LUBRICATING     OILS 

I'AGE 

Composition,  Treatment  and  Examination  of  Petroleums 
and  their  chief  Distillation  Products  —  Classification 
and  Examination  of  Lubricating  Oils  —  Fatty  Oils, 
etc.,  in  Mineral  Oils  .  .  .  .  .  .  153 

CHAPTER   VI 

MILK    AND    BUTTER 

Composition  and  Properties  of  Milk  and  Butter  —  Examin- 
ation of  Milk,  Cream  and  Skim  Milk  —  Adulteration  of 
Milk  —  Sterilising  and  Pasteurising  of  Milk—  Bacterial 
Contents  of  Milk  —  Analysis  of  Condensed  Milk  — 
Analysis  of  Butter  and  Margarine  ....  182 

CHAPTER   VII 

STARCH  AND  ITS  DECOMPOSITION  PRODUCTS—  FLOUR, 
BARLEY  AND  MALT 

Properties  and  Identification  of  Starch  —  Estimation  of 
Starch  in  Flour,  Grain,  etc.  —  Estimation  of  Alcohol 
—  Production,  Composition  and  Diastatic  Value  of 
Malt  —  Classification  and  Examination  of  Flour  — 
Examination  of  Barley  and  Malt  ....  232 

^> 

CHAPTER   VIII 

PRESERVATIVES    AND    ARTIFICIAL    COLOURING    MATTERS 
IN    FOODS 

Preservatives  in  Foods,  general  —  Preservatives  in  Milk, 
Cream,  Butter  and  Margarine  —  Preservatives  in 
Meat,  Wine,  Beer,  Preserves,  Fruit  Juices,  etc.  — 
Artificial  Colouring  Matters  in  Foods  .  .  .  .  280 


INDEX      .  .         .  .         .    -     .         .         .     319 


INDUSTRIAL 
ORGANIC      ANALYSIS 


INTRODUCTION. 

THIS  chapter  contains  some  general  remarks  on  the 
work  to  be  described,  and  suggestions  for  the  use  of  this 
volume. 

Sampling. — A  few  words  may  first  be  said  concerning 
the  operation  of  taking  the  sample  for  analysis. 
Although  the  student  may  not,  in  the  ordinary  course  of 
events,  have  many  opportunities  of  taking  his  own 
samples  direct  from  large  masses  of  material,  yet  he 
should  never  lose  sight  of  the  importance  of  always  work- 
ing on  a  fair  average  sample  of  the  consignment  or  batch 
of  material  to  be  valuated  or  tested  by  chemical  and 
physical  examination.  It  is  obvious  that  if  the  sample 
for  analysis  has  been  picked  by  a  prejudiced  person,  or 
carelessly  taken,  the  analytical  results  obtained  may,  in 
many  cases,  fall  far  short  of  representing  the  average 
properties  or  composition  of  the  bulk  of  the  material ; 
and  the  analysis,  however  carefully  and  efficiently  it 
may  have  been  carried  out,  may  be  practically  valueless, 
at  any  rate  to  the  person  who  would  be  most  directly 
interested  in  obtaining  reliable  information  as  to  the 
quality  of  the  material. 

If  unable  to  sample  from  the  bulk,  the  student  should, 
at  all  events,  always  make  a  point  of  taking  a  genuine 

I.O.A.  I 


2          Industrial  Organic  Analysis 

average  sample  for  analysis  from  the  material  supplied 
to  him  ;  he  should  mention  in  his  report  the  weight  or 
bulk  of  the  sample  received  for  analysis,  the  date 
received  in  the  laboratory,  the  date  analysed,  and  as  far 
as  is  known  to  him,  the  age  and  previous  history  of  the 
sample  on  receipt.  The  latter  points  are  mainly  of 
importance  when  dealing  with  materials  which  are  liable 
to  decomposition  on  keeping. 

The  method  of  sampling  will  naturally  depend  on  the 
nature  of  the  substance  dealt  with.      An  example  of 
a   method  for  sampling   solid   substances   is   given   in 
Chapter  I.  (Sampling  of  Coal)  ;    the  details  given  will 
serve  as  an  illustration  of  the  methods  which  may  be 
employed  for  the  sampling  of  hard  solids,  though  various 
modifications  may  naturally  suggest  themselves  when 
dealing  with  other  materials  of  a  similar  nature.     Soft 
or  semi-solid  material,  such  as  butter,  soap,  solid  fats, 
anthracene,  etc.,  may  be  sampled  by  means  of  the  auger- 
shaped  instrument  described  in  Chapter  VI.,  p.   227, 
several  samples  being  taken  in  different  directions  from 
the  barrel,  drum  or  box,  and,  if  possible,  melted  together, 
the  analytical  sample  being  taken  from  the  well-mixed, 
molten  material ;   an  example  of  this  method  is  given  in 
Chapter   VI.    (Sampling  of  Butter).     This   method   of 
melting  together  several  samples  taken  from  the   bulk 
should  always  be  adopted  if  the  material  can  be  melted 
at  a  conveniently  low  temperature,  and  without  decom- 
position or  loss  of  volatile  constituents.     The  sampling 
of  liquids  usually  presents  the  least  difficulty,  though 
care   should   always   be   taken   that   any   sediment   or 
immiscible  liquid  which  may  be  present  be  stirred  into 
the  mass  before  taking  the  sample.     (See  Chapter  VI., 
p.  190.) 
Accuracy  of  Results. — Many  of  the  analytical    pro- 


Introduction  3 

cesses  described  are  only  intended  to  yield  results 
accurate  to  two,  or,  at  the  most,  three  figures,  as  in 
industrial  work  extreme  accuracy  may  often  be  sacri- 
ficed with  advantage  for  rapidity  and  ease  in  working. 
In  many  cases,  indeed,  highly  exact  methods  of  analysis, 
even  if  available,  would  be  too  lengthy  and  complicated 
for  frequent  use  in  the  technical  laboratory,  while  very 
little  would  be  gained  by  obtaining  results  correct  to 
more  than,  say,  one-tenth,  or  even  i  per  cent.  For 
these  reasons,  time  may  often  be  saved  by  carrying  out 
weighings  on  a  balance  weighing  correctly  to  the  nearest 
5  or  10  milligrams,  instead  of  on  the  accurate  analytical 
balance,  while  approximate  volumetric  methods  may 
often  be  made  to  take  the  place  of  the  more  exact 
gravimetric  methods.  Even  in  fairly  accurate  work,  the 
rough  balance  may  be  used  when  weighing  comparatively 
large  quantities,  say  over  10  grams  of  material. 

The  student  will  learn  to  appreciate  the  relative 
accuracy  aimed  at  in  the  various  processes  described,  and 
to  modify  his  methods  of  working  accordingly  ;  for 
example,  in  an  operation  such  as  the  determination  of 
the  iodine  value  of  fats,  where  scientifically  accurate 
results  are  desirable,  and  the  amount  of  material  weighed 
is  small,  it  will  be  necessary  to  use  the  accurate  analytical 
balance.  On  the  other  hand,  in  operations  such  as  the 
distillation  tests  for  coal  tar  and  petroleum  or  similar 
products,  the  material  and  its  distillation  fractions  may 
be  measured  with  sufficient  accuracy  by  means  of 
graduated  cylinders,  as  the  methods  of  analysis  employed 
in  such  cases  are  only  capable  of  yielding  approximately 
accurate  results.  To  sum  up,  the  degree  of  accuracy 
required  in  the  purely  quantitative  operations  will 
depend,  firstly,  on  the  degree  of  accuracy  of  the  results 
which  the  method  is  capable  of  yielding,  and  secondly,  on 

I — 2 


4          Industrial  Organic  Analysis 

the  amount  of  material  dealt  with  ;  in  this  connection, 
however,  it  may  be  pointed  out  that  although  a  method 
of  analysis  may  only  be  capable  of  yielding  approximate 
results,  it  may  still  be  of  the  greatest  importance  to 
adhere  closely  to  the  prescribed  experimental  conditions 
in  carrying  it  out. 

The  results  of  an  analysis  should  never  be  expressed  as 
being  accurate  to  a  greater  degree  than  is  warranted  by 
the  methods  by  which  they  have  been  obtained  ;  thus 
in  some  cases  it  is  well  to  express  percentages  of  con- 
stituents as  falling  within  fairly  wide  limits  (see,  for 
example,  the  estimation  of  cotton  seed  oil  in  arachis  oil 
(Chapter  III.,  p.  no)). 

Recording  of  Work  and  Reporting  of  Results. — The  usual 
notebook  account  should  include  not  only  a  record  of 
work  done  and  observations  made  in  its  course,  but  also 
reasons*  for  undertaking  each  operation  described, 
explanations  of  all  observations  of  interest,  conclusions 
to  be  drawn  from  such  observations  and  the  analytical 
results,  and  explanations  for  all  conclusions  drawn.  In 
addition  to  such  an  account,  which  will  deal  chiefly  with 
the  scientific  aspect  of  the  work,  and  may  be  regarded 
mainly  as  a  private  record  for  future  guidance,  it  is 
suggested  that  the  student  should  endeavour  to  write 
reports  dealing  primarily  with  the  purely  technical  or 
commercial  aspect  of  the  results  obtained.  Such  reports 
should  be  written  so  as  to  be  perfectly  intelligible  to 
persons  having  little  or  no  theoretical  knowledge  of 
Physics  or  Chemistry  ;  they  should  be  brief  and  logical, 
containing  no  statements  which  cannot  easily  be  justified. 
Doubtful  or  speculative  conclusions  should  either  be 
omitted,  or,  at  any  rate,  definitely  characterised  as  such. 
Although  technical  terms  cannot  well  be  entirely  avoided, 
the  meaning  of  such  terms  as  might  be  unfamiliar  to  the 


Introduction  5 

person  to  whom  the  report  is  addressed,  together  with 
their  qualitative  or  quantitative  significance,  should  be 
clearly  and  briefly  explained  in  the  discussion  of  results. 
Any  such  explanations  should  be  based  mainly  on 
observed  or  known  facts,  chosen  and  stated  so  as  to 
corroborate  the  conclusions  drawn  from  the  results  of  the 
investigation  ;  they  should  be  in  plain  language,  and  as 
free  as  possible  from  theoretical  explanations. 

The  report  should  commence  with  the  particulars 
relating  to  the  sample  or  samples,  referred  to  above. 
Next,  note  may  be  made  of  any  obvious  properties  of  the 
samples,  such  as  colour,  smell,  taste,  consistency,  etc., 
which  may  appear  to  be  of  interest.  The  results  of  the 
investigation  may  now  be  stated,  after  which  will  follow 
a  brief  discussion  of  the  technical  or  commercial  signifi- 
cance of  the  analytical  results  and  other  observations  of 
interest,  and  finally  the  conclusions  which  may  be  drawn 
as  to  the  technical  or  commercial  value  of  the  samples. 

Suggestions  for  the  Use  of  this  Volume. — It  is  hardly 
expected  that  the  student  will  find  time  to  work  through 
all  the  examples  given  in  this  volume  ;  a  selection  may 
be  made,  partly  according  to  the  individual  interest  of 
the  student,  and  partly  with  the  object  of  obtaining 
experience  of  as  many  different  types  of  analytical 
methods  as  possible. 

It  is  suggested  that  the  material  for  examination 
should  generally  be  selected  or  prepared  by  the  teacher, 
especially  when  the  problem  is  to  detect  impurities  or 
adulterants.  In  such  cases  it  may  often  be  advisable  for 
the  teacher  to  prepare  samples  containing  known 
amounts  of  impurities  or  adulterants  from  genuine 
products,  and  to  give  the  stude'nt  an  opportunity  of 
making  parallel  tests  or  analyses  on  material  which  is 
known  to  be  genuine  and  unadulterated  ;  for  example, 


6          Industrial  Organic  Analysis 

two  samples  may  be  supplied,  with  the  information  that 
one  of  them  is  suspected  to  be  adulterated,  the  problem 
being  to  detect  whether  this  be  the  case,  and  if  so,  the 
nature  and  approximate  extent  of  the  adulteration,  as 
far  as  can  be  ascertained  by  the  methods  available  to  the 
student.  In  this  way,  the  student  will  be  able  to 
test  thoroughly  the  methods  which  he  employs,  and  to 
form  an  opinion  as  to  their  relative  value,  and  the 
significance  of  the  results  obtained  by  their  means. 
This  method  of  making  parallel  tests  on  perfectly 
genuine  material  should  always  be  employed  in  trying 
new  methods,  or  in  doubtful  cases.  In  preparing 
adulterated  material  for  analysis,  the  foreign  con- 
stituents should  be  added  in  such  amounts  as  may  be 
detected  without  great  difficulty  by  the  methods  avail- 
able to  the  student,  as  the  object  is  rather  to  give  him 
an  opportunity  of  familiarising  himself  with  the  methods 
and  principles  of  industrial  analysis  than  to  test  his  skill 
as  an  analyst  by  confronting  him  with  problems  which 
might  even  tax  the  resources  of  the  expert. 

Another  form  of  problem  which  may  be  presented  to 
the  student  is  to  supply  him  with  a  sample  of  material 
for  examination,  and  to  ask  him  either  to  report  generally 
on  its  technical  or  commercial  value,  or  to  report  on  its 
value  for  some  particular  purpose  ;  a  problem  of  this 
nature  may  admit  of  wider  possibilities  than  that  indi- 
cated above,  which  simply  demands  the  detection  of 
certain  foreign  materials.  It  may  also  require  more 
extensive  inspection  of  standard  works  of  reference  on 
the  subjects  under  investigation,  and  is  at  all  times  a  test 
of  the  student's  business  capacities. 

(All  temperatures  are  given  in  degrees  Centigrade,  and 
all  polarimetric  readings  in  angular  degrees,  except  where 
otherwise  indicated.) 


CHAPTER   I 

COAL   AND   COKE 

INTRODUCTORY. 

COAL,  which  is  at  present  the  most  important  of  the 
sources  of  energy  available  for  industrial  purposes,  has 
resulted  from  the  mineralisation  of  the  wood  of  pre- 
historic plants,  which  has  remained  buried  in  the  crust 
of  the  earth,  and  undergone  changes  as  the  result  of 
which  it  has  more  or  less  completely  lost  its  organic 
nature.  In  availing  ourselves  of  the  latent  energy  stored 
up  in  the  coal,  by  burning  it  in  our  grates  and  furnaces, 
we  are,  broadly  speaking,  reversing  the  processes  by 
which  the  plant  made  use  of  the  radiant  energy  of  the 
sun  in  carrying  out  the  endothermic  changes  necessary 
to  convert  water,  carbonic  acid  and  simple  mineral  sub- 
stances into  complex  vegetable  tissues.  In  view  of 
the  complex  nature  of  the  process  of  mineralisation  of 
wood,  which  has  resulted  in  the  formation  of  coal,  as  well 
as  the  difference  in  the  conditions  under  which  it  must 
have  taken  place  in  different  localities,  it  is  hardly  sur- 
prising that  we  should  meet  with  an  almost  endless 
variety  of  coals,  passing  by  almost  imperceptible  grada- 
tions, from  the  comparatively  soft  brown  coals,  or 
lignites,  which  have  still  preserved  the  fibrous  structure 
of  the  wood,  to  the  hard,  black,  shiny  anthracites,  which 
show  conchoidal  fracture,  and  display  but  little  evidence 
of  their  organic  origin. 

Composition  and  Classification  of  Coals. — The  chief 
uses  of  coal  are  as  follows  :  (i)  steam  raising  ;  (2)  house- 
hold use  ;  (3)  metallurgical  and  other  manufacturing 
operations,  and  for  smiths'  forges  ;  (4)  the  production 
of  coke,  gas,  tar  and  ammonia. 


8          Industrial  Organic  Analysis 

Coals  are  broadly  classified  according  to  their  technical 
applications,  which  may  be  determined  from  the  results 
of  chemical  and  physical  examination.  The  interpreta- 
tion of  the  results  of  a  chemical  analysis  of  coal  generally 
falls  under  two  headings.  Firstly,  there  are  to  be  con- 
sidered such  constituents  as  water,  sulphur  and  mineral 
matter,  which  must  be  regarded  as  more  or  less  adven- 
titious impurities,  and  which  dilute  the  coal,  reduce  its 
calorific  value  or,  if  present  in  considerable  amounts,  may 
render  it  unsuitable  for  certain  purposes.  Secondly,  the 
results  of  an  analysis  enable  one  to  classify  the  coal,  and 
to  determine  approximately  the  uses  for  which  it  is  best 
suited.  The  soundest  method  for  the  classification  of 
coals  according  to  their  chemical  composition  has  been 
shown  to  be  that  which  is  primarily  based  or  the  per- 
centages of  carbon  and  hydrogen  calculated  on  the  coal, 
less  water,  sulphur  and  ash  ;  thus,  if  S  =  the  sum  of  the 
percentages  of  the  water,  sulphur  and  ash  found  in  the 
coal,  and  C  =  the  percentage  of  carbon  found  in  the  coal, 
then  the  percentage  of  carbon  on  the  coal,  less  water, 
sulphur  and  ash,  will  be  given  by 
P  -IPO 
"  100  -  S 

The  reason  for  eliminating  these  constituents  from  the 
calculation  is  that  they  may  vary  considerably  in  amount 
in  coals  of  essentially  the  same  type,  and  thus  affect  the 
percentages  of  carbon  and  hydrogen,  as  determined  on 
the  coal  itself,  to  such  an  extent  that  no  true  comparison 
can  be  made. 

The  accompanying  table  shows  how  coals  are  broadly 
classified  into  five  main  groups,  on  this  basis.  The 
lignites  and  the  true  anthracites  are  omitted  from  the 
scheme. 

Apart  from  the  content  of  more  or  less  adventitious 
impurities,  such  as  water,  sulphur  and  ash,  the  main 
points  which  come  into  consideration  in  judging  of  the 
value  of  a  coal  for  any  particular  purpose  are  as 
follows  : — 

The  amount  of  gas  and  other  volatile  combustible 
matter  formed  on  heating ;  this  determines  the 


Coal  and  Coke 


l>  -*  v 

131 


o 
bbg 


GO 


10 

oc~ 


o 

CO 


10 

M 


§ 

PO 


ro 


u 


O 

* 


OO 


r 


^ 

OH 


>! 

X 

O 


2         y> 

A         T 


U 
E 


10 

o 

O                  u 

0                 0 

^J- 

in 

m 

1-            ^t- 

1 

1 

10 

co 

ii 

n             u 

o            m 

to 

m 

V 

o             u 

"5                     "+• 

0 

m 

C 

JN                       H 

H             .        CO 

CO 

T 

OC 

5                         C 

TN                 ON 

TO 

o 

T 

)                   O 

^ 

oo 

GO                 CO                  ON 

M 

<M 

co             ^f             in 

io        Industrial  Organic  Analysis 

suitability  of  the  coal  for  gas  or  tar  manufacture,  and  the 
length  of  flame  on  combustion.  The  semi-anthracitic 
and  anthracitic  coals,  which  burn  either  with  a  short 
flame  or  practically  no  flame  at  all  (Groups  4  and  5  in 
the  above  table),  are  more  difficult  to  kindle  than  the 
long  flame  coals,  but  give  a  more  intense  heat,  and 
produce  little  or  no  smoke.  In  these  groups  are  included 
the  well-known  Welsh  steam  coals  which  are  so  largely 
used  with  marine  boilers.  Their  freedom  from  smoke 
also  renders  them  suitable  for  drying  and  curing  hops  and 
malt,  and  for  horticultural  purposes.  Generally  speak- 
ing, there  is  a  direct  relationship  between  the  percentage 
of  hydrogen  and  the  length  of  flame,  which  both  decrease 
in  passing  from  Group  2  to  Group  5. 

Some  coals  (Groups  2,  3  and  4)  become  fused  or  semi- 
fused  on  heating  or  burning ;  these  are  referred  to  as 
caking  coals.  The  non-caking  coals,  which  are  included 
in  Groups  i  and  5,  do  not  soften  in  this  way,  so  that  the 
individual  lumps  remain  separate  during  combustion  and 
provide  for  a  freer  access  of  air  than  do  the  caking  coals. 
These  peculiarities,  as  well  as  the  relative  length  of  the 
flame,  must  be  taken  into  account  in  considering  the 
type  of  furnace  for  which  the  coal  is  best  suited.  The 
caking  power  also  influences  the  nature  of  the  coke  ; 
thus,  it  will  be  noticed  that  the  non-caking  coals  of 
Groups  i  and  5  produce  pulverulent  cokes,  while  those 
of  Groups  2,  3  and  4  all  produce  more  or  less  coherent 
cokes.  The  condition  of  the  coke  is  further  affected 
by  the  amount  of  gas  formed  on  heating  or  combustion, 
which,  generally  speaking,  varies  directly  with  the  per- 
centage of  hydrogen.  We  accordingly  find  that  in 
passing  from  Group  2  to  Group  4,  the  coke  becomes 
denser  and  less  porous. 

The  non-caking  coals  of  Group  i  differ  from  those  of 
Group  5  in  burning  with  a  long  flame.  They  are  mainly 
used  for  steam  raising  and  are  also  suitable  for  gas 
making  in  some  cases. 

The  coals  belonging  to  Groups  2  and  3  are  sometimes 
referred  to  as  the  "  bituminous  coals,"  though  it  should 
be  understood  that  they  have  no  connection  with  the 
substance  known  as  bitumen.  The  softer  varieties, 


Coal  and  Coke  1 1 

which  are  rich  in  volatile  matter,  are  included  in  Group  2, 
and  are  usually  known  as  the  cannel  coals  ;  they  are 
generally  used  for  making  gas,  tar  and  ammonia.  The 
coals  of  Group  3,  which  yield  more  coke  than  those  of 
the  former  group,  are  used  in  smiths'  forges  and  for 
manufacturing  coke,  Gas,  tar  and  ammonia  are  some- 
times also  obtained  from  these  coals. 

The  true  anthracites  and  the  lignites,  which  have  not 
been  dealt  with  hitherto,  take  their  places  at  either  end 
of  the  scale  in  the  above  classification.  The  former 
contain,  as  a  rule,  over  93  per  cent,  of  carbon,  and  less 
hydrogen  than  the  anthracitic  coals  belonging  to 
Group  5  ;  they  burn  without  flame,  are  shiny  jet-black 
in  appearance,  show  conchoidal  fracture,  and  sometimes 
have  a  specific  gravity  as  high  as  1-6.  The  lignites,  on 
the  other  hand,  contain  less  carbon  and  more  hydrogen 
than  the  coals  of  Group  i  ;  they  are  brownish-black  in 
appearance,  comparatively  soft,  and  show  a  fibrous 
structure  ;  owing  to  the  relatively  large  amount  of 
water  and  mineral  matter  which  they  contain,  they  have 
a  lower  calorific  value  than  the  true  coals.  Commercially 
they  are  of  considerably  less  importance  than  the 
latter. 

The  relations  between  the  chemical  composition  of 
coals  and  their  chief  properties  and  uses  will  be  treated 
of  in  greater  detail  when  the  methods  for  determining 
the  various  .constituents  are  described. 

The  above  classification,  and  the  explanatory  remarks 
which  follow  it,  must  not  be  interpreted  in  too  literal 
a  sense  ;  it  is  very  difficult  to  place  all  coals  in  clearly 
defined  groups  forming  a  simple  system  of  classification, 
while  their  uses,  though  primarily  dependent  on  their 
chemical  composition,  must,  in  many  cases,  largely  be 
determined  by  commercial  considerations,  such  as  the 
cost  of  the  coal  itself,  cost  of  transport,  and  the  value 
of  the  products  obtainable  from  it  (i.e.,  gas,  tar,  ammonia 
or  coke),  which  latter  can  only  be  satisfactorily  deter- 
mined by  trials  on  a  manufacturing  scale.  It  is,  however, 
generally  possible  to  form  an  idea  of  the  uses  to  which 
a  coal  may  be  put  with  greatest  advantage  from  the 
results  of  a  chemical  examination ;  and  moreover,  if  a 


12        Industrial  Organic  Analysis 

certain  kind  of  coal  has  been  found  to  be  suitable  for  a 
certain  purpose,  chemical  examination  will  show  whether 
subsequent  supplies  are  likely  to  fulfil  the  same  conditions 
as  the  original  sample.  Again,  it  will  always  be  useful 
to  have  a  check  on  the  water,  ash  and  sulphur,  which, 
if  present  in  large  amounts,  invariably  reduce  the  value 
of  the  coal  as  fuel.  It  should  always  be  borne  in  mind 
that  no  analysis  of  coal  will  be  of  much  value  unless 
carried  out  on  a  fair  average  sample.  (See  below  under 
"  Sampling.") 

The  accompanying  table  (p.  14)  shows  the  results  of 
analyses,  collected  from  various  sources  (most  from 
"  Analyses  of  British  Coals  and  Cokes  "),  of  typical  coals 
and  cokes,  and  their  principal  uses  ;  together  with  the 
classification  and  explanatory  remarks  given  above,  it 
may  be  used  as  a  guide  for  interpreting  the  results 
obtained  by  the  analytical  methods  described  below. 

CHEMICAL  EXAMINATION  OF  COAL  AND  COKE 

Sampling. — In  all  technical  and  commercial  analytical 
work,  the  operation  of  taking  the  sample  for  analysis 
is  of  the  utmost  importance  ;  in  the  case  of  coal  and  coke, 
where  such  large  masses  are  dealt  with,  this  remark 
applies  with  especial  force  ;  it  is  obvious  that  a  carefully 
picked  sample  may  differ  very  considerably  in  composi- 
tion from  a  sample  which  has  been  taken  systematically 
and  without  prejudice,  with  the  object  of  ascertaining, 
as  far  as  possible,  the  average  composition  of  the  entire 
mass.  Although  the  sampling  of  coal  and  coke  can 
hardly  be  described  as  a  laboratory  operation,  directions 
for  systematic  sampling  are  given  here,  owing  to  the 
extreme  importance  of  this  operation  to  the  analyst. 
Whether  or  not  the  student  has  had  the  opportunity  of 
taking  his  sample  from  a  large  mass  of  material,  he 
should,  at  all  events,  follow  out  the  latter  part  of  the 
following  directions  in  dealing  with  the  reduced  sample. 

Quantities  of  coal  or  coke,  as  the  case  may  be,  amount- 
ing in  all  to  2  to  3  cwts.,  are  taken  at  equal  intervals 


Coal  and  Coke  13 

during  unloading,  or  from  all  sides  as  well  as  the  interior 
of  the  heaps,  and  removed  to  a  place  where  the  material 
will  be  protected  from  rain,  dampness  or  direct  sunlight. 
The  whole  is  broken  up  into  pieces  about  the  size  of 
apples,  well  mixed  and  arranged  in  the  form  of  a  square 
layer  about  20  cm.  deep.  Diagonals  are  marked  out 
on  this  square,  and  the  material  in  two  opposite  triangles 
is  removed  ;  the  remainder  is  then  rearranged  into  a 
square,  and  the  process  repeated  until  about  200  Ibs. 
remain,  which  are  broken  into  pieces  about  the  size  of 
nuts,  and  systematically  reduced,  by  the  process  just 
described,  to  about  10  Ibs.  The  sample  thus  obtained 
is  preserved  in  air-tight  metal  boxes  till  required  for 
use,  when  it  is  further  reduced  as  follows  : — The  whole 
is  well  mixed,  whereupon  500  grams  are  abstracted  in 
such  a  way  as  to  obtain  an  average  sample,  and  reduced 
to  a  fine  powder,  preferably  in  a  small  hand  mill,  so 
constructed  as  to  avoid  loss  of  dust  during  the  grinding. 
The  powder  is  placed  in  a  dry  bottle  provided  with  a 
well- fit  ting  glass  stopper,  and  thoroughly  mixed  by 
shaking.  Samples  for  analysis  are  taken  from  this  bottle 
as  required. 

Hygroscopic  Water,  (a)  In  Coal. — 2  to  4  grams  of  the 
powdered  coal  are  weighed  between  a  pair  of  well-ground 
watch  glasses,  so  that  moisture  shall  not  be  attracted 
through  undue  exposure  to  the  air,  and  dried  at  a 
temperature  of  105°  on  the  shelf  of  an  air  oven,  till 
constant  in  weight  ;  the  time  required  is  usually  about 
two  hours.  Overheating  of  the  sample  must  be  avoided, 
or  appreciable  quantities  of  volatile  matter  other  than 
water  may  be  lost ;  the  process  is  always  liable  to  a 
certain  error  due  to  the  cause  just  mentioned,  as  well  as 
oxidation  of  the  coal,  which  occurs  to  a  slight  extent 
when  the  latter  is  heated  in  contact  with  air  to  the 


14        Industrial  Organic  Analysis 


C,A-5                                0 

CO 

i^<  ^       Is 

«n 

Tj- 

CO 

o 

M 

CO 

Tj- 

rj- 

in 

O 

^"gh"!     ^-8> 

be  v  j  |5 

N 

CO 

* 

0 

m 

in 

m 

•n 

in 

>n 

|l|5       § 

ON 

^ 

in 

'      0 

^ 

0 

CO 

cc 

lo*    1 

rf 

ON 

CO 

(S 

O 

ON 

oo 

CO 

8 

O 
00 

o 
oo 

in 

00 

fmi 

1 

1 

9 

00 

V) 

op 

M 

ON 

rj- 

£ 

op 

CO 

1 

jy  </>       M"  £?.?•£ 

<N 

liiill§ 

1 

1 

o 

cb 

0 

b 

S" 

6 

M 

ON 

CO 

M 

ON 

M 

Tf 

Tf 

>o~s*  ^8, 

CO 

N 

<N 

CO 

CO 

co 

|| 

- 

0 

CO 

N 

CO 

r}- 

N 

9 

H 

o 

^& 

< 

M 

M 

CO 

M 

M 

M 

M 

<s 

<N 

CO 

IM 

^ 

M 

op 

9 

CO 

Tf 

ON 

CO 

9 

ft 

CO 

J4M 

§ 

6 

O 

M 

6 

oo 
6 

in 

6 

CO 

ON 

6 

ON 

6 

1 

g  rfta*S 

CO 

o 

N 

CO 

ON 

O 

M 

1" 

o 

g*M 

^ 

0 

M 

M 

M 

M 

M 

. 

fenffc^ 

in 

o 

N 

O 

oo 

oo 

ON 

0 

N 

N 

^S>^8 

<s 

CO 

* 

M 

CO 

in 

CO 

0 

O 

** 

SflVH^ 

^ 

CO 

0 

m 

t^ 

9 

N 

M 

M 

CO 

|^ft§ 

N 

CO 

Tf 

* 

* 

in 

in 

in 

in 

in 

II 

in 

rt- 

^ 

ON 

CO 

^ 

in 

^ 

IN 

w 

.   & 

a 

CO 

Tj- 

CO 

oo 

oo 

00 

00 

H 

00 

1 

'd    • 

bJO   • 

_ 

Cfl 

H 

d 

o 

o 

. 

6 

9 

rC 

iS 

a 

:ription  of  Coal  or  Coke 

te,  Pennsylvania 

BO 

of 

tuo 
E 

8 
1 

manufacturing 
x>al,  Leicester  . 

If 

aanufacturing,  co 
use  coal,  Lancashi 

Durham  . 

Durham  (J.  Patti 

al,  Durham 

,  Commentry,  Fr 
r)  . 

I 

Anthraci 

Anthraci 

Welsh  st 

n 

C/2 

eg 

S   . 

o  '"""" 
u 

Steam,  n 
and  ho 

Gas  coal, 

Gas  coal, 

o 
o 

§ 

$ 

11 

O    CO 

o 

Coal  and  Coke 


CO 

CO                   CO           Tj- 

M 

to                iO         u"N 

H 

O                      M             t> 

oo 

T+-              M          ON 

b 

oo           oo       r^ 

oo 

1           6       1 

1 

°ON                  t^ 

O         u 

1        0 

1 

M                      d 

CO                    CO 

M              C 

)      6 

? 

ON                   CO              M 

C 
10                    M              CO            C 

o 

)             ^N 

rt        ^ 

CO                    MM 

b          6      6      c 

)         O 

•S   1 

ON            co         O 

M                      CO            M 

VO                      t^           M              k 

ob           V      f^      c 

M 

^         co 
*        M 

S     §      en 

>^        r^H        i—4 

1           6      6 

M            in      N      a 
6          6      M      c 

)      6 

*  8  "§ 

M                      <S            00 
M                      MM 

?          1        1 

1 

S  ^j    O 
u    g   ,— 

^j       O       CU 

i>*          oj       o^ 

Tf- 

£?     c     cj 

ob           6       <s 

«             1          1 

1 

^03 
tut)  T3  ^y 

O                if)         ro 

5,     I    I 

1 

JH    s    0 

-3  3   g    . 

x  ^   S,  g 

5 

11            -^i 

QJ         CO          O         ^ 

rj-               O         *O 

CO              00           Tj-        vj 

0         >0 

rf  2  "o  a 

&             00        00 

C"^         *o       c*** 

10          "^t*      o 

VO                   ON          ON         C 

^       r>» 

0         00 

<U       0       ^     o 

tj      4J     ^       CU 

g    g  'a  3 

a)            bo   • 

^     •         f&      •         ^r^ 

9              . 

ctf     c^ 

i  i 

03              C              O 
-<->              rt              en 

^    rS     c     5 

4-f         "         O         ^ 

*  •  :|  •  i 

~                       ^2                      ,C 

$*->                             en 

N                                   T3 

•*•    1  •    I 

3        c/5        p 
9  '  '  '  jf        *" 

(jit 

P 

JS  JS  -2    e 

0     0     rt     0 

>  >  o  ^ 

rt            S  IH        ° 

bo    '       ,2             cj 

CD" 

S            IS     ^ 

"O  In       £ 

LH                      <1   17           ri4 

o  >      J     8    ' 

s  1 

•^      ^2 

f     p                       flj   ^Q 

_M    rf          "*^ 

u      ^ 

-|    |o    § 

|^      8^     H     c 
^^     sofe:     «     ) 

3      a 

a  s 

fl    Cti           0)  '-f          £ 

CM        en  (u 
o3  ^         30        rt 

co       ffi       55 

CD   be       C  ^        — 
CO       '43     •        CD 

§«    «G   3    • 

O           g           O      ( 

o       iT 

«  ^ 

o      o 
J      0 

16        Industrial  Organic  Analysis 

temperature  employed  for  expelling  the  water.  In  most 
cases,  however,  the  results  thus  obtained  will  be  found  to 
be  sufficiently  accurate  for  industrial  purposes,  though 
if  the  sample  is  subsequently  to  be  analysed  for  hydrogen, 
it  is  advisable  to  employ  a  more  accurate  method,  for 
errors  in  the  figure  for  water  as  moisture  will  affect  the 
figure  for  hydrogen  as  determined  by  combustion. 

Where  greater  accuracy  is  desired,  the  water  as  mois- 
ture should  be  determined  as  follows  : — 2  to  4  grams  of 
powdered  coal  are  spread  in  a  thin  layer  on  a  clock  glass 
and  placed  over  sulphuric  acid  in  a  vacuum  desiccator 
which  is  then  evacuated.  The  sample  is  weighed  every 
24  hours  till  no  further  loss  in  weight  takes  place ; 
before  and  during  the  weighing,  the  coal  should  be 
covered  by  a  clock  glass  and  exposed  to  the  air  as 
little  as  possible.  If  a  good  vacuum  is  maintained  in  the 
desiccator,  the  drying  should  be  complete  in  24  to  48 
hours. 

The  proportion  of  hygroscopic  water  present  in  coals 
is  very  variable  ;  excluding  the  lignites,  which  may 
contain  up  to  30  per  cent.,  coals,  especially  when  freshly 
raised,  may  contain  as  much  as  15  per  cent,  of  water, 
though  this  limit  is  seldom  exceeded.  Broadly  speaking, 
a  good  coal  should  not  contain  more  than  3  to  4  per  cent, 
of  water,  while  in  a  good  anthracite  the  percentage  of 
water  should  lie  between  nil  and  1*5,  or  at  most,  2. 

Hygroscopic  Water,  (b)  In  Coke. — According  to  Arth, 
100  to  200  grams  of  coke,  broken  into  small  lumps,  are 
weighed  into  a  porcelain  dish  and  dried  in  an  air  oven  at 
150  to  160°,  until  constant  in  weight.  Samples  should 
be  kept  in  air-tight  vessels,  as  they  easily  lose  their 
water  on  exposure  to  air. 

Hygroscopic  water  not  only  dilutes  the  fuel,  but 
also  reduces  its  calorific  value  owing  to  the  heat  which  it 


Coal  and   Coke  17 

absorbs  on  evaporation.  It  is  sometimes  the  custom  to 
specify  a  certain  limit  for  the  amount  of  hygroscopic 
water  in  coal,  and  to  make  a  proportionate  deduction  in 
the  price  paid  if  this  limit  is  exceeded. 

Ash. — The  residue  of  mineral  matter  left  on  complete 
combustion  of  the  coal  or  coke  is  weighed  as  ash.  The 
amount  of  ash  thus  found  does  not  represent  the  total 
amount  of  mineral  matter  originally  present  in  the  fuel, 
and,  moreover,  varies  somewhat  with  the  mode  of  com- 
bustion, depending,  among  other  things,  on  the  amount 
of  sulphur  left  in  the  residue  and  the  amount  of  carbon 
dioxide  lost  by  the  alkaline  earth  carbonates.  The 
determination  of  the  ash  should  be  carried  out  separately, 
and  not  combined  with  the  elementary  analysis  or  the 
determination  of  volatile  combustible  matter  and  coke. 
The  following  directions  for  the  determination  are  due 
to  Arth  : — I  to  5  grams  of  the  finely  powdered  coal  (or 
coke)  are  weighed  off  in  a  platinum  dish  or,  failing  this,  a 
porcelain  dish,  about  7  cm.  in  diameter  ;  at  first  a  gentle 
heat  is  applied,  in  order  to  avoid  caking  and  subsequent 
difficulty  in  combustion,  the  dish  being  covered  with  a 
piece  of  platinum  foil  so  long  as  there  is  any  danger  of 
loss  by  decrepitation.  The  subsequent  heating  is  best 
carried  out  in  a  muffle  furnace,  the  dish  being  placed  on 
a  piece  of  platinum  foil,  to  avoid  contact  with  the 
siliceous  material  of  the  furnace.  The  temperature  is 
gradually  raised  to  a  bright  red  heat ;  if  the  layer  of  ash 
is  thick  it  should  be  cautiously  stirred  from  time  to  time 
with  a  stout  piece  of  platinum  wire.  When  the  com- 
bustion is  judged  to  be  complete,  the  dish  and  its  con- 
tents are  cooled  in  a  desiccator  and  weighed.  Complete 
combustion  is  effected  with  difficulty  if  graphitic  matter 
is  present  in  the  residue ;  unburnt  carbon  should  be 
tested  for  in  the  cooled  residue  by  adding  a  few  drops  of 

I.O.A.  2 


i8        Industrial  Organic  Analysis 

alcohol,  when  the  carbon  particles  will,  if  present,  be 
observed  floating  on  the  liquid.  In  this  case  the  alcohol 
should  be  removed  by  evaporation  and  the  combustion 
of  the  residue  completed  over  the  blow-pipe  flame.  In 
difficult  cases,  the  combustion  may  also  be  facilitated  by 
allowing  a  gentle  current  of  oxygen  to  impinge  on  the 
residue  while  it  is  being  heated,  care,  of  course,  being 
taken  that  particles  of  ash  are  not  blown  away. 

The  proportion  of  ash  is  greatest  in  coals  which  are 
rich  in  volatile  combustible  matter ;  in  these  it  may 
sometimes  exceed  15  or  even  20  per  cent.  As  in  the  case 
of  the  water,  a  certain  maximum  limit  for  the  ash  in  coal 
might  be  specified  by  the  purchaser,  say  8  or  10  per  cent., 
and  a  deduction  made  in  the  price  paid  if  the  limit  were 
exceeded.  The  undesirability  of  a  large  proportion  of 
ash  in  coal  is  obvious  ;  the  ash  dilutes  the  fuel,  tends  to 
obstruct  the  draught  during  combustion,  and  carries 
with  it  some  of  the  heat  of  the  fire,  as  well  as  unburnt 
fuel,  on  falling  from  the  grate.  Generally  speaking,  it 
may  be  said  that  a  good  coal  should  yield  less  than  6  per 
cent,  of  ash.  The  same  applies  to  coke,  especially  when 
used  for  metallurgical  or  similar  operations,  in  which  the 
chemical  action  of  the  ash  has  to  be  considered. 

Volatile  Combustible  Matter,  Combustible  Residue  and 
Coke. — When  coal  is  heated  out  of  contact  with  air,  part 
of  the  carbon  is  lost,  mainly  in  the  form  of  volatile  hydro- 
carbons and  carbon  monoxide ;  hydrogen,  oxygen, 
water  and  ammonia  are  given  off  at  the  same  time. 
The  non-volatile  residue,  which  consists  of  an  impure 
carbon  containing  the  greater  part  of  the  mineral  matter 
of  the  original  coal,  is  known  as  coke.  The  total  volatile 
matter,  less  the  hygroscopic  water,  is  determined  as 
"  volatile  combustible  matter  "  ;  the  non- volatile  resi- 
due, less  the  ash,  is  known  as  the  "  non- volatile  com- 
bustible residue,"  the  "  fixed  combustible  residue,"  or 
the  "  fixed  carbon."  The  latter  term  should  be  avoided, 
as  it  is  somewhat  misleading. 

The  importance  of  the  yield  and  nature  of  the  coke  has 
already  been  pointed  out ;  the  yield  and  nature  of  the 


Coal  and  Coke  19 

volatile  combustible  matter  are  of  importance  if  the  coal 
is  to  be  used  for  gas  or  tar  manufacture,  and  the  yield  of 
volatile  combustible  matter,  as  determined  by  a  labora- 
tory operation,  is  a  useful  guide  in  classifying  the  coal 
under  examination.  Bearing  in  mind  that  the  amount 
as  well  as  the  chemical  composition  of  the  volatile  matter 
obtained  from  a  coal  vary  to  some  extent  with  the 
conditions  under  which  the  destructive  distillation  is 
carried  out,  it  follows,  firstly,  that  the  results  of  any 
laboratory  operation  should  not  be  interpreted  too 
literally  in  judging  of  the  probable  behaviour  of  a  coal 
when  submitted  to  destructive  distillation  on  a  manu- 
facturing scale,  especially  as  regards  the  nature  of  the 
distillate,  and,  secondly,  that  the  estimation  of  the 
volatile  and  non-volatile  matter  should  be  carried  out, 
as  far  as  possible,  under  certain  standard  conditions  in 
order  that  the  results  obtained  may  be  truly  comparable. 
The  determination  described  should  not  be  looked  on  as 
an  exact  analytical  operation  but  as  a  "  test  "  or  "  assay," 
yielding  results  which,  if  not  strictly  accurate,  never- 
theless furnish  useful  indications.  The  following  method 
for  estimating  the  volatile  and  non-volatile  matter  in 
coal  is  that  adopted  by  the  American  Coal  Committee.1 

Place  i  gram  of  the  undried  fresh  coal  in  a  platinum 
crucible  weighing  20  to  30  grams,  and  provided  with  a 
tightly  fitting  lid.  Support  the  crucible  on  a  platinum 
triangle,  6  to  8  cm.  above  the  top  of  a  Bunsen  burner, 
which  should  be  burning  with  a  flame  20  cm.  high  when 
free.  Heat  for  exactly  7  minutes  in  a  place  free  from 
draughts.  The  upper  surface  of  the  crucible  lid  should 
burn  clear,  but  the  under  surface  should  remain  covered 
with  carbon.  The  volatile  combustible  matter  should 
be  calculated  on  the  dry  coal.  Duplicate  analyses 
should  agree  within  0-5  per  cent.  The  coke  and  non- 
volatile combustible  residue  may  be  estimated  from  the 

1  Journ.  Amer.  Chem.  Soc.,  Vol.  XXI.,  pp.  1122  to  1126. 

2 — 2 


2O        Industrial  Organic  Analysis 

weight   of  the   residue  in   the  crucible,   the  latter   by 
deducting  the  ash  as  previously  determined. 

The  accompanying  table  shows  the  percentages  of 
volatile  matter  yielded  by  the  different  varieties  of  coal, 
anthracite  and  lignite,  as  determined  by  Mahler  : — 


Description  of  Coal. 

Volatile  combustible  matter 
calculated  on  coal,  less  ash. 

Anthracites    . 

3  to  5 

Semi-anthracitic 

6  to  14 

Bituminous  for  steam  and 

coking 

15  to  25 

Bituminous  for  blast  fur- 

naces and  forges 

25  to  30 

Cannel  .... 

About  50 

Lignites  of  good  quality  . 

About  50 

The  outstanding  feature  to  be  noticed  here  is  the 
gradual  diminution  in  the  volatile  matter  in  passing  from 
the  cannels  to  the  anthracites.  The  above  groups,  from 
the  cannels  to  the  semi-anthracitic  coals,  inclusive, 
roughly  correspond  with  the  Groups  i  to  5,  respectively, 
in  the  table  on  p.  9. 

The  association  of  a  high  percentage  of  volatile  matter 
with  a  high  percentage  of  ash  is  sometimes  explained  by 
the  assumption  that  the  mineral  matter  has  enclosed  the 
mother  substance  of  the  coal,  and  thus  hindered  its 
decomposition  by  the  escape  of  gas. 

Coke  should,  in  view  of  its  method  of  production,  only 
contain  slight  quantities  of  volatile  matter.  If  desired, 
this  may  be  determined  as  described  for  coal. 

Sulphur. — This  constituent  is  usually  present  in  coal 
in  the  form  of  iron  pyrites  and  calcium  sulphate.  When 
the  coal  is  submitted  to  destructive  distillation  part  of 
the  sulphur  is  driven  off,  chiefly  in  the  form  of  organic 
sulphur  compounds,  owing  to  the  decomposition  of  the 
pyrites.  The  determination  of  sulphur  in  coal  and  coke 


Coal  and  Coke  21 

is  often  carried  out  by  the  Eschka-Fresenius  method,  as 
follows  : — 

i  gram  of  the  finely  powdered  material  is  mixed  in  a 
platinum  crucible  with  twice  its  bulk  of  a  mixture  of 
i  part  of  dry  sodium  carbonate  and  2  parts  of  calcined 
magnesia  ;  the  uncovered  crucible  is  placed  in  an  inclined 
position,  and  its  lower  half  is  heated  to  a  low  red  heat  till 
the  colour  of  the  contents  changes  from  a  grey  to  a  yellow 
or  brown  ;  it  is  then  allowed  to  cool,  the  ash  mixed  with 
one  half  to  i  gram  of  ammonium  nitrate,  and  heated 
again  to  redness  for  5  to  10  minutes,  the  crucible  this 
time  being  covered  with  a  lid.  After  cooling,  the 
crucible  is  placed  in  a  beaker  of  water,  and  the  con- 
tents detached  by  heating  the  liquid.  The  crucible  is 
removed,  washed  carefully  with  water  in  order  to 
remove  adhering  solution,  and  the  washings  added 
to  the  contents  of  the  beaker.  During  the  process 
of  ignition  just  described,  the  sulphur  contained  in  the 
coal  or  coke  will  have  been  oxidised  to  soluble  sulphates, 
which  may  be  determined  in  the  solution  by  precipitating 
with  barium  chloride  and  weighing  the  precipitate,  in  the 
usual  manner  described  in  all  text-books  on  quantitative 
inorganic  analysis. 

The  limits  of  variation  of  the  proportion  of  sulphur  in 
coal  are  usually  from  one  half  to  2  per  cent.,  while  in 
anthracites  it  may  only  be  present  in  the  smallest  traces. 
Generally  speaking,  it  may  be  said  that  a  good  coal  or 
coke  should  contain  less  than  i  per  cent,  of  sulphur.  It 
is  important  that  coke  for  use  in  the  blast  furnace  should 
contain  a  minimum  of  this  constituent  owing  to  its 
undesirable  influence  on  the  properties  of  the  iron. 

Nitrogen. — This  constituent  exists  in  the  coal  in  the 
combined  form,  and  originates  from  the  proteins  present 
in  the  original  wood.  When  coal  is  submitted  to  destruc- 
tive distillation,  the  nitrogen  escapes  mainly  in  the  form 


22        Industrial  Organic  Analysis 

of  ammonia,  and  also  as  aromatic  amines,  pyridine 
bases,  etc.  The  nitrogen  in  coal  is  conveniently  esti- 
mated by  the  method  of  Kjeldahl,  which  is  very  com- 
monly used  for  the  estimation  of  nitrogen  in  foodstuffs, 
fertilisers  and  other  materials.  The  method  depends  on 
the  destruction  of  the  nitrogen  radicles  in  organic  com- 
pounds and  their  conversion  into  ammonium  sulphate, 
by  the  action  of  boiling  concentrated  sulphuric  acid  in 
the  presence  of  a  little  copper  or  mercury  sulphate. 
When  the  digestion  with  the  acid  is  complete,  the 
mixture  is  made  alkaline,  and  the  ammonia  distilled  off 
into  a  known  volume  of  standard  acid  and  estimated  by 
titration.  If  it  is  desired  to  include  the  nitrogen  as 
nitrates  and  nitrites  in  the  estimation,  the  method 
must  be  modified  as  subsequently  described. 

.  To  carry  out  the  determination,  i  gram  of  the  powdered 
coal  is  introduced  into  a  Kjeldahl  digestion  flask  together 
with  20  c.c.  of  pure  sulphuric  acid  of  specific  gravity  1-84, 
and  i  gram  of  granular  copper  oxide.  The  further 
addition  of  10  grams  of  potassium  sulphate  will  have  the 
effect  of  shortening  the  process  of  digestion  owing  to  the 
higher  temperature  attained  through  the  presence  of 
potassium  bisulphate.  The  Kjeldahl  digestion  flask  (see 
Fig.  i)  should  be  made  in  Jena  or  other  resistant  glass, 
and  should  have  a  capacity  of  about  200  c.c.  It  is  pear 
shaped,  and  has  a  neck  about  16  cm.  long  and  2  cm.  in 
diameter.  Placing  the  charged  flask  in  an  inclined 
position  on  a  piece  of  wire  gauze,  heat  is  carefully  applied 
until  all  frothing  has  ceased  ;  the  mixture  may  then  be 
boiled  briskly  until  it  has  become  perfectly  clear  and 
only  retains  the  bright  blue-green  colour  of  the  copper 
salt.  At  this  stage,  the  flask  is  removed  from  the  flame 
and  held  upright,  while  powdered  potassium  perman- 
ganate is  dropped  into  the  hot  mixture  in  very  small 
quantities  at  a  time,  until,  after  shaking,  a  greenish 
purple  colour  remains  ;  in  this  way  the  complete  oxida- 


Coal  and  Coke  23 

tion  of  all  organic  matter  is  ensured.  After  cooling 
water  is  added  with  extreme  caution,  small  quantities  at 
a  time  being  allowed  to  flow  down  the  side  of  the  flask, 
while  the  contents  are  continually  being  agitated;  the 
whole  is  transferred  to  a  700  c.c.  flask,  the  total  water 


FIG.  L — Kjeldahl  Digestion  Flasks  on  Stand. 

added,  including  rinsings,  amounting  to  200  to  300  c.c. 
A  few  pieces  of  zinc  foil  are  added  to  secure  even  boiling, 
and  an  excess  of  concentrated  sodium  hydroxide  solution 
(about  100  c.c.  of  a  saturated  solution)  is  poured  down 
the  side  of  the  flask  in  such  a  way  as  to  avoid,  as  far  as 
possible,  its  mixing  with  the  acid  liquid.  Immediately 
after  the  addition  of  the  alkali,  the  flask  is  connected  up 


24        Industrial  Organic  Analysis 

by    means    of    a    rubber    stopper    to    the    distillation 
apparatus. 

The  accompanying  illustration  (Fig.  2)  will  give  an  idea 
as  to  the  general  arrangement  of  the  apparatus.     The 


FIG.  2. — Distillation  Apparatus  for  KjeldahPs  Process  for  the 
Determination  of  Nitrogen. 

receiving  flask  contains  20  c.c.  of  decinormal  sulphuric 
acid,  and  the  tube,  which  dips  just  below  the  surface 
of  the  acid  liquid,  is  provided  with  a  bulb  in  order  to 
prevent  suck-back  into  the  distillation  flask.  For  this 
purpose  an  ordinary  pipette  may  be  used.  If  preferred, 


Coal  and  Coke  25 

the  vapour  may  be  condensed  in  a  straight  tube  con- 
denser, in  which  case  the  condenser  tube  should  be 
brought  near  the  surface  of  the  standard  acid,  not 
dipping  into  the  latter.  If  only  a  few  analyses  are  to  be 
made,  ordinary  Liebig  condensers,  fitted  with  adapters, 
may  be  used  ;  if  many  analyses  are  to  be  made,  it  is 
more  convenient  to  employ  a  series  of  metal  tubes 


FIG.  3. — Distillation  Apparatus   for  Kjeldahl's  Process  for  the 
Determination  of  Nitrogen. 

placed  vertically  in  a  metal  tank  through  which  water 
may  be  made  to  circulate,  as  shown  in  Fig.  3.  Arrange- 
ments such  as  this,  which  may  be  obtained  from  most 
dealers  in  chemical  apparatus,  admit  of  a  number  of 
distillations  being  carried  out  simultaneously.  What- 


26        Industrial  Organic  Analysis 

ever  the  arrangement  used,  the  distillation  flask  should 
be  connected  with  a  bulb-trap  for  preventing  alkaline 
spray  from  being  carried  over  into  the  receiver.  When 
the  liquid  has  been  boiled  nearly  to  dryness,  all  the 
ammonia  is  certain  to  have  been  distilled  over  into  the 
acid,  which  is  then  titrated  with  decinormal  sodium 
hydroxide  or  baryta  solution  in  order  to  determine  the 
amount  of  ammonia  absorbed.  The  indicator  used  may 
be  litmus,  methyl  orange  or  cochineal,  but  not  phenol 
phthalein. 

It  is  generally  advisable  to  perform  the  analysis  in 
duplicate  and  also  to  do  a  blank  test  in  order  to  determine 
the  amount  of  ammonia  in  the  materials  used,  which 
should,  however,  be  negligible.  If  the  analysis  is  carried 
out  for  the  first  time,  a  parallel  determination  should 
be  carried  out  with  a  pure  organic  substance  containing 
nitrogeri,  such  as  hippuric  acid. 

Modification  of  Kjeldahl's  method  to  include  Nitrogen 
of  Nitrates. — It  is  obvious  that  working  according  to  the 
method  just  described,  the  nitrogen  of  nitrates  will  be 
lost  in  the  form  of  nitric  acid.  If  it  is  desired  to  include 
the  nitrogen  present  in  this  form,  the  process  must  be 
modified  as  follows  : — 

Gunning's  Method:  The  sample  is  well  mixed  with 
35  c.c.  of  pure  sulphuric  acid  in  which  has  been  dissolved 
3  per  cent,  of  salicylic  acid,  and  the  whole  is  shaken  at 
frequent  intervals  for  10  minutes.  5  grams  of  sodium 
thiosulphate  and  10  grams  of  potassium  sulphate  are 
then  added,  and  the  mixture  is  carefully  heated  until 
frothing  has  ceased,  after  which  it  is  further  treated  as 
directed  above. 

The  above  modification  is  especially  applicable  in  the 
analysis  of  fertilisers,  in  which  a  considerable  proportion 
of  the  nitrogen  may  be  present  in  the  form  of  nitrates. 


Coal  and  Coke  27 

Further  instances  of  the  application  of  the   Kjeldahl 
process  will  be  given  in  Chapters  VI.  and  VII. 

The  proportion  of  nitrogen  in  coal  may  vary  from 
0-2  to  2  per  cent.  The  determination  of  this  constituent 
is  chiefly  of  interest  as  furnishing  a  rough  estimate  of 
ammonia  and  other  nitrogenous  products  obtainable  on 
distillation  of  the  coal.  If  this  question  does  not  come 
into  consideration,  the  estimation  is  often  omitted,  and 
the  nitrogen  is  grouped  together  with  the  oxygen,  in  the 
statement  of  the  results  of  the  analysis,  both  being 
determined  by  difference. 

Carbon  and  Hydrogen. — The  importance  of  knowing 
the  percentages  of  carbon  and  hydrogen  for  the  purpose 
of  classifying  the  coal  has  already  been  pointed  out  in 
the  introductory  section  of  this  chapter.  The  determina- 
tion of  the  calorific  value  of  the  coal,  as  described  below, 
is  also  based  on  the  elementary  analysis. 

The  method  employed  for  the  determination  of  carbon 
and  hydrogen  in  coal  is  essentially  the  same  as  that 
employed  for  determining  these  constituents  in  organic 
compounds  which  contain  nitrogen  and  sulphur.  The 
details  of  the  operation  will  be  found  described  in  many 
books  dealing  with  practical  organic  chemistry,  and 
will  therefore  not  be  repeated  here  ;  it  will  only  be 
necessary  to  call  attention  to  a  few  points  of  interest  in 
the  present  case. 

About  half  a  gram  of  the  finely  powdered  coal  should 
be  weighed  off  in  a  porcelain  boat,  about  7  cm.  long  and 
7  to  8  mm.  broad,  and  then  mixed  with  about  twice  its 
bulk  of  dry,  finely  powdered  copper  oxide,  great  care 
being  taken  to  avoid  loss.  The  combustion  tube  should 
be  packed  throughout  with  lead  chromate,  in  the  usual 
way,  in  order  that  all  sulphur  may  be  retained  as  lead 
sulphate  and  prevented  from  escaping  as  sulphur 
dioxide,  a  short  reduced  copper  spiral  being  placed  in 


28        Industrial  Organic  Analysis 

the  rear  end,  in  order  to  decompose  nitrogen  oxides. 
While  volatile  matter  is  being  distilled  from  the  coal, 
oxygen  should  only  be  allowed  to  pass  through  the  tube 
at  the  rate  of  one  bubble  in  two  seconds  ;  the  speed  may 
be  doubled  when  only  coke  is  left.  The  lead  chromate 
should  not  be  heated  so  strongly  that  it  fuses  with  the 
glass  of  the  tube  ;  moreover,  lead  sulphate,  which  will 
be  formed  by  the  interaction  of  the  chromate  and 
sulphur  of  the  coal,  is  not  quite  stable  at  such  high 
temperatures. 

In  calculating  the  percentage  of  hydrogen  from  the 
water  absorbed  by  the  drying  tube,  allowance  must  be 
made  for  the  hygroscopic  water,  accurately  determined 
as  described  above,  this  being  deducted  from  the  total 
water  weighed,  in  order  to  arrive  at  the  water  formed 
by  combustion  of  the  hydrogen.  As  has  already  been 
pointed  out,  powdered  coal  is  hygroscopic,  so  that  unless 
the  sample  for  analysis  is  preserved  in  a  well  stoppered 
bottle  during  the  interval  between  the  water  determina- 
tion and  the  elementary  analysis,  serious  errors  may  be 
made.  The  re-calculation  of  the  percentages  of  carbon 
and  hydrogen  on  the  coal,  less  water,  ash  and  sulphur, 
has  been  described  on  p.  8.  The  method  for  calculating 
the  calorific  value  of  the  coal  from  the  results  of  the 
elementary  analysis  is  described  below.' 

Oxygen. — There  is  no  direct  method  available  for 
determining  this  constituent ;  it  is  therefore  determined 
by  difference,  deducting  the  sum  of  the  percentages  of 
the  carbon,  hydrogen,  hygroscopic  water,  ash,  sulphur 
and  nitrogen  from  one  hundred.  The  figure  thus 
obtained  will  naturally  be  affected  by  the  accumulated 
errors  of  the  previous  determinations ;  the  oxygen 
content  is,  for  this  and  other  reasons,  no  longer  used  as  a 
basis  for  the  classification  of  coals.  If  the  nitrogen  has 
not  been  estimated,  both  this  and  the  oxygen  will  be 
determined  together,  by  difference. 

The  amount  of  hydrogen  found  in  excess  of  that 
required  to  combine  with  the  oxygen  in  the  coal  (exclud- 


Coal  and  Coke  29 

ing,  of  course,  the  oxygen  of  the  hygroscopic  water)  to 
form  water,  is  known  as  the  "  disposable  hydrogen." 
Generally  speaking,  the  disposable  hydrogen  may  be 
taken  as  a  relative  measure  of  the  amount  of  gas  obtain- 
able from  the  coal,  and  an  indication  of  the  length  of 
flame  which  will  be  produced  on  combustion.  These 
relationships  will  be  apparent  on  referring  to  the  tables 
given  above. 

Determination  of  the  Calorific  Value  of  Coal  and  Coke. — 
The  amount  of  heat  theoretically  obtainable  from  a  given 
weight  of  coal  or  coke  determines  to  a  great  extent  its 
value  as  a  fuel  ;  it  is  often  the  custom  to  base  the 
commercial  valuation  of  fuels  on  the  results  of  determina- 
tions of  calorific  value,  which  have  been  carried  out  on 
fair  average  samples. 

The  calorific  value  of  a  fuel  may  be  stated  in  several 
ways : — 

(a)  As  the  number  of  calories  produced  by  burning 
i  kilo,  of  the  fuel,  the  calorie  being  the  amount  of  heat 
required  to   raise   the  temperature  of  I  kilo,  of  water 
from  4°  to  5°  C. 

(b)  As  the  number  of  British  Thermal  Units  obtainable 
from  i  Ib.  of  the  fuel,  the  B.T.U.  representing  the  amount 
of  heat  required  to  raise  the  temperature  of  i  Ib.  of 
water  through  i°  F.  at  39-1°  F. 

(c)  As  the  number  of  pounds  of  water  which  may  be 
evaporated  at  the  boiling   point  by  the  combustion  of 
i  Ib.  of  the  fuel. 

Calculation  of  Calorific  Value  from  the  Results  of 
Elementary  Analysis. — This  method  is  chosen  for  descrip- 
tion in  detail,  as  it  is  based  on  the  results  of  processes 
which  have  already  been  described.  The  calculation 
is  based  on  Dulong's  assumption  that  the  amount  of  heat 
given  out  on  the  burning  of  a  fuel  is  equal  to  the  sum  of 
the  amounts  of  heat  produced  on  the  combustion  of  its 
separate  elements,  the  whole  of  the  oxygen  present  being 
considered  as  already  combined  with  sufficient  of  the 
hydrogen  present  in  the  form  of  water.  Knowing  the 
percentages  of  the  carbon,  disposable  hydrogen  and 
sulphur  present  in  the  coal,  and  the  calorific  values  of 
these  elements,  it  should  be  possible  to  calculate  the 


30        Industrial  Organic  Analysis 

calorific  value  of  the  fuel  itself.  The  above  assumption 
is,  of  course,  not  theoretically  justifiable,  but  experience 
has  shown  that  when  applied  in  the  case  of  coal,  the 
results  obtained  are  sufficiently  accurate  for  practical 
purposes.  The  present  method  cannot  be  applied  to 
liquid  fuels,  as  here  the  heats  of  formation  of  the  con- 
stituent compounds  are  too  considerable  to  be  left  out 
of  account. 

The  following  data  are  employed  in  the  calculation  : — 

Calories. 

I  kilo,  of  Carbon  in  burning  to  CO2  produces  8,140 
i        ,,       Sulphur  ,,  SO2        ,,        2,160 

i        „       Hydrogen        „  H2O        „       34>5oo 

as  water 

i        ,,       Hydrogen        ,,  H2O        ,,       28,900 

as  steam 

In  practice,  it  is  usualto  take  the  calorific  value  of 
hydrogen  as  34,500  calories  per  kilo.  Then  if 

H  be  the  percentage  of  Hydrogen  in  the  Fuel, 
C  ,,  ,,  Carbon 

S  ,,  ,,  Sulphur 

and  O  ,,  ,,  Oxygen 

the  quantity  of  heat,  Q  calories,  obtainable  from 
i  kilo,  of  the  fuel  is  usually  calculated  from  Dulong's 
formula  :— 

34,500  (H  —  i  O)  +  8,140  C  -f  2,160  S 
100 

Representative  results  of  calorific  determinations  on 
the  various  kinds  of  coal  are  included  in  the  table  on 
p.  9.  As  a  rule,  the  results  obtained  for  coal  by  the 
method  described  do  not  differ  from  those  obtained  by 
the  calorimetric  method  by  more  than  about  3  per  cent.  ; 
in  many  cases  the  variation  is  less  than  i  per  cent. 
Mahler  has  shown  that  the  variations  in  the  results 
obtained  by  the  two  methods  are  considerable  in  the  Case 
of  lignites,  peat,  and  especially  mineral  oils.  The 
calorific  values  of  the  latter  fuels  cannot  be  accurately 


Coal  and  Coke  31 

calculated  from  the  carbon  and  hydrogen  content,  but 
must  be  determined  by  the  calorimetric  method,  in 
which  a  known  weight  of  the  fuel  is  burnt  in  a  closed 
bomb  containing  oxygen  under  pressure,  and  the  heat 
evolved  calculated  from  the  rise  in  temperature  indicated 
by  a  thermometer  in  the  water  contained  in  the  calori- 
meter in  which  the  bomb  is  placed.  This  method,  which 
is  often  used  for  the  determination  of  the  calorific  value 
of  coal,  is  fully  described  in  most  of  the  works  mentioned 
at  the  end  of  this  chapter. 

The  heat  actually  available  in  practice  naturally  falls 
short  of  the  amount  of  heat  theoretically  obtainable  on 
combustion  of  the  fuel  ;  owing  to  loss  of  heat  with  the 
chimney  gases,  with  the  ash  as  it  falls  from  the  grate, 
radiation  and  incomplete  combustion,  the  available  heat 
is  usually  less  than  80  per  cent,  of  that  theoretically 
obtainable.  The  results  of  laboratory  determinations  of 
calorific  value  are,  however,  none  the  less  valuable,  as 
they  enable  the  analyst  to  estimate  the  relative  heat 
values  of  different  fuels  01  kinds  of  coal  when  used  under 
approximately  similar  conditions. 

PHYSICAL  EXAMINATION  OF  COAL  AND  COKE. 

In  the  table  on  p.  9  will  be  found  the  specific 
gravities  of  the  different  varieties  of  coal ;  it  will  be 
noticed  that  there  is  a  gradual  increase  in  the  specific 
gravity  in  passing  from  the  varieties  which  contain  least 
carbon  to  the  anthracitic  group.  The  determination  of 
this  constant  may  usually  be  omitted,  as  surer  indica- 
tions of  the  nature  of  the  coal  are  obtained  from  a 
chemical  examination. 

It  is  sometimes  of  importance  to  determine  the 
porosity  and  the  crushing  strength  of  coke  ;  the  latter  is 
of  importance  when  the  coke  is  to  be  used  in  the  blast 
furnace,  as  it  should  be  capable  of  bearing  the  weight  of 
the  material  above  it  without  being  crushed  to  powder, 
in  which  case  it  would  tend  to  obstruct  the  draught.  As 
has  been  pointed  out  previously,  the  nature  of  a  coke 
depends,  not  only  on  the  nature  of  the  coal  from  which 
it  has  been  prepared,  but  also  on  the  method  of  prepara- 


32        Industrial  Organic  Analysis 

tion  ;  thus,  the  coke  produced  in  coking  ovens  is,  as  a 
rule,  denser  and  of  greater  mechanical  strength  than  that 
which  is  obtained  as  a  by-product  from  gas  and  tar 
manufacture  in  the  gas  retort.  The  methods  for 
determining  the  porosity  and  crushing  strength  of  coke 
are  fully  described  in  Stillmann's  "  Engineering 
Chemistry." 

REFERENCES. 

"  Engineering  Chemistry,"  by  Stillmann.  1910.  Williams  and 
Norgate,  London,  and  the  Chemical  Publishing  Co.,  Easton,  Pa. 

"  Technical  Methods  of  Chemical  Analysis,"  Vol.  I.,  Part  I., 
by  Lunge,  translated  by  Keane.  1911.  Gurney  and  Jackson, 
London. 

"  Contributions  a  1'Etude  des  Combustibles,"  by  Mahler. 
1903.  Librairie  Polytechnique,  Paris. 

"  Chemisch-Technische  Analyse,"  by  Post.  1908.  Fr.  Vieweg 
und  Sohn,  Brunswick. 

"  Analyses  of  British  Coals  and  Cokes,-"  by  Greenwell  and 
Elsden.  1907.  Chichester  Press,  London. 

"  Gas  and  Fuel  Analysis  for  Engineers,"  by  Gill.  1911.  Wiley 
and  Sons,  New  York. 

"  An  Introduction  to  the  Study  of  Fuel,"  by  Brislee.  1912. 
Constable  and  Co.,  London. 

"  Fuel  and  Refractory  Material,"  by  Humboldt  Sexton.  1909, 
Blackie  and  Son,  London. 


CHAPTER  II 

COAL  TAR  AND  ITS  DISTILLATION  PRODUCTS 
INTRODUCTORY. 

WHEN  coal  is  submitted  to  dry  distillation  it  yields 
three  main  products,  viz.,  coal  gas,  water  containing 
ammonia  in  solution,  and  coal  tar.  The  present  chapter 
deals  with  the  examination  of  the  latter  substance  and 
its  chief  distillation  products,  with  special  reference  to 
their  most  important  application,  i.e.,  the  production  of 
materials  which  form  the  starting  points  in  the  manu- 
facture of  dyes,  disinfectants  and  other  valuable  sub- 
stances, derived  chiefly  from  the  aromatic  hydrocarbons. 
Coal  tar  is  chiefly  obtained  as  a  by-product,  either  in  the 
manufacture  of  coke,  in  which  case  it  is  known  as  coke- 
oven  tar,  or  in  the  manufacture  of  coal  gas,  when  it  is 
known  as  gas  tar.  Blast  furnace  and  generator  gas  tars, 
which  are  sometimes  obtained  as  by-products  in  the 
manufacture  of  pig  iron  and  producer  gas,  respectively, 
have  not  the  same  commercial  importance  as  the  two 
first  mentioned  products. 

Properties  and  Composition  of  Coal  Tar.— Crude  coal 
tar  is  a  dark  brown  or  black,  more  or  less  viscous  fluid, 
with  a  creosote-like  smell.  It  is  usually  heavier  than 
water,  the  tars  having  specific  gravities  under  i-ooo  being 
useless  for  the  production  of  dyes.  The  latter  point  will 
be  more  fully  dealt  with  below. 

From  what  has  been  said  in  the  previous  chapter,  it 
will  be  gathered  that  the  proportion  of  tar  yielded  by 
different  kinds  of  coal  varies  considerably  ;  the  nature  of 
the  tar  is  influenced,  not  only  by  the  quality  of  the  coal 
from  which  it  is  distilled,  but  also  to  a  considerable 
extent  by  the  temperature  at  which  the  distillation  is 
carried  out.  Generally,  when  low  temperatures  are 

I.O.A.  3 


34        Industrial  Organic  Analysis 

employed,  the  resulting  tars  consist  mainly  of  liquid  and 
solid  paraffins,  olefines  and  the  more  complex  phenols. 
Higher  temperatures,  on  the  other  hand,  tend  to  give 
rise  to  the  formation  of  aromatic  hydrocarbons,  free 
carbon,  and  phenol  rather  than  homologues  of  phenol ; 
olefines  and  acetylenes  are  formed  in  smaller  quantities, 
while  the  paraffins  practically  disappear.  It  may  also 
be  mentioned  that  larger  amounts  of  gas  are  formed  at 
high  than  at  low  temperatures.  These  differences  are 
explained  fully  in  Lunge's  "  Coal  Tar  and  Ammonia  " 
(Vol.  I.,  pp.  16  et  seq.),  where  it  is  pointed  out  that  the 
action  of  heat  tends  to  encourage  molecular  condensa- 
tions with  the  formation  of  substances  such  as  naphtha- 
lene and  anthracene,  and  the  elimination  of  hydrogen, 
either  in  the  elementary  state,  or  in  the  form  of  highly 
hydrogenated  hydrocarbons  such  as  methane.  The 
tendency  towards  the  formation  of  phenol  rather  than 
the  cresols  at  higher  temperatures  is  similarly  explained, 
while  the  formation  of  free  carbon,  either  in  a  finely 
divided  state  or  as  a  graphitic  mass,  may  be  regarded  as 
the  last  stage  resulting  from  the  tendency  towards  mole- 
cular condensation  brought  into  play  by  the  action  of 
heat. 

The  temperature  employed  in  coke  ovens  is  generally 
lower  than  in  gas  retorts ;  hence  gas  tar  is  generally 
richer  in  aromatic  hydrocarbons  than  coke-oven  tar. 
The  majority  of  coke  oven  tars  are,  however,  produced 
in  such  a  way  as  to  be  suitable  for  the  manufacture  of 
dyestuffs.  The  most  valuable  tars  are  those  from  which 
the  aromatic  hydrocarbons,  notably  benzene,  toluene 
and  anthracene,  may  be  conveniently  isolated  in  a  state 
of  sufficient  purity  for  the  production  of  colours ;  for  this 
it  is  essential  that  the  tar  should  contain  a  minimum  of 
paraffins,  as  these  cannot  conveniently  be  separated  on 
a  manufacturing  scale,  while  their  presence  in  apprecia- 
able  quantities  may  give  rise  to  considerable  difficulty 
in  the  nitration  and  other  chemical  processes.  Tars 
containing  comparatively  large  amounts  of  open  chain 
hydrocarbons  are  of  no  great  value  ;  they  are  chiefly 
used  for  the  production  of  burning  and  lubricating  oils 
and  solid  paraffins. 


Coal  Tar  and  its  Distillation  Products    35 

Coal  tar,  as  produced  for  the  manufacture  of  colours, 
contains  the  following  substances  as  its  most  important 
constituents  :— 

Water  and  ammonia. 

Benzene,  toluene,  xylenes  and  higher  homologues  of 
benzene. 

Phenol,  cresols,  naphthols  and  other  phenolic  sub- 
stances of  higher  molecular  weight. 

Amine  bases  such  as  aniline  and  its  homologues. 

Naphthalene,  anthracene,  phenanthrene  and  their 
homologues,  as  well  as  other  hydrocarbons  containing 
condensed  benzene  nuclei,  of  high  molecular  weight, 
such  as  pyrene  and  chrysene. 

Pyridine,  quinohne  and  bases  of  a  similar  nature. 

Nitriles,  carbazoles,  etc. 

Sulphur  containing  compounds  of  which  the  chief  are 
carbon  disulphide  and  thiophen. 

In  order  to  isolate  the  chief  constituents,  the  coal  tar 
is  submitted  to  a  preliminary  distillation  in  retorts  placed 
over  an  open  fire.  The  following  fractions  are 
collected  : — 

(1)  First  runnings,  up  to  about  120°  C. 

(2)  Light  oils  or  crude  naphtha,  up  to  about  170°  C. 

(3)  Middle  or  carbolic  oils,  from  170°  to  230°  C. 

(4)  Creosote  oil,  from  230°  to  270°  C. 

(5)  Anthracene  oil,  over  270°  C. 

The  first  two  fractions  are  often  collected  together, 
in  which  case  they  will  be  referred  to  as  the  "  total  light 
oils  "  ;  they  are  accompanied  by  the  aqueous  ammonia 
which  remained  in  the  crude  tar.  The  still  residue 
consists  of  pitch. 

The  composition,  treatment  on  a  manufacturing  scale, 
and  especially  the  laboratory  examination  of  these 
fractions  and  the  products  obtained  from  them,  will  be 
dealt  with  below  under  separate  headings.  The  accom- 
panying table  contains  figures  taken  from  Lunge's 
'  Technical  Methods  of  Chemical  Analysis,"  showing  the 
results  of  the  preliminary  distillation  of  four  typical  tars 
distilled  by  different  methods  for  the  production  of 

3—2 


36        Industrial  Organic  Analysis 

aromatic  compounds :  it  will  be  noticed  that '  the 
ordinary  gas  tar  yields  more  light  oils  and  less  anthracene 
oils  than  the  chamber  retort  gas  tars  and  the  coke  oven 
tars  ;  further  differences  will  be  pointed  out  under  the 
descriptions  of  the  determinations  of  specific  gravity  and 
free  carbon. 


Same  Coal  in  Gas 

— 

Retorts. 

Chamber 
Retort, 
Gas  Tar. 

Coke  Oven 
(Otto),  Tar. 

Vertical. 

Horizontal. 

Ammoniacal  water 

2-17 

3-50 

_ 

2-69 

Total  light  oil      . 

5-85 

3-IO 

0*40 

1-38 

Middle,  or  carbolic  oil 

12-32 

7-68 

l     IO'2O  I 

3'46 

Heavy,  or  creosote  oil 

H'95 

IOT5 

I     I0  2°  I 

9'93 

Anthracene  oil    . 

15-96 

n'54 

30-10 

24-76 

Pitch 

4975 

62-00 

53  '90 

56-44 

Naphthalene 

— 

4-70 

— 

Free  carbon        .:.•• 

II'IO 

THE  EXAMINATION  OF  COAL  TAR. 

Under  this  heading  the  determination  of  specific 
gravity  and  free  carbon  and  a  distillation  test  are 
described. 

(a)  Specific  Gravity. — The  water  is  first  separated  as 
follows  :  the  tar  is  allowed  to  stand  in  a  conical  flask 
immersed  in  water  at  50°,  for  24  hours ;  the  water 
collecting  at  the  top  is  poured  off  as  completely  as 
possible,  the  rest  being  absorbed  by  drawing  a  piece  of 
filter  paper  over  the  surface.  The  tar  is  then  cooled  to 
15°  and  its  specific  gravity  determined  at  this  tempera- 
ture. For  this  purpose,  the  pyknometer  or  hydrometer 
cannot  well  be  used  owing  to  the  viscosity  of  the  tar  at 
ordinary  temperatures.  Lunge  advises  the  use  of  an 
ordinary  cylindrical  weighing  bottle  of  about  50  c.c. 
capacity,  a  vertical  groove,  about  2  mm.  wide  and  2  mm. 


Coal  Tar  and  its  Distillation  Products    37 

deep,  being  cut  in  the  glass  stopper.  The  bottle  is 
weighed  dry  and  empty,  and  then,  when  filled  with  water 
at  15°,  in  the  usual  way.  It  is  then  about  two-thirds 
filled  with  the  tar  and  placed,  without  the  stopper,  in 
hot  water  for  one  hour  to  get  rid  of  air  bubbles.  After 
cooling,  the  bottle  and  stopper  are  weighed  with  the 
tar.  Water  is  then  added  to  fill  the  bottle,  and  the 
whole  weighed  after  allowing  to  stand  in  water  at  15°. 
From  these  data,  the  weight  and  volume  of  the  tar  are 
readily  calculated,  and  hence  the  specific  gravity. 

As  previously  mentioned,  the  specific  gravity  of  tars 
which  consist  mainly  of  aromatic  hydrocarbons  is 
usually  above  i-ooo  ;  if  under  this  figure,  the  tar  is 
generally  not  worth  any  further  treatment.  The  mean 
specific  gravity  of  gas  tar  from  horizontal  or  sloping 
retorts  is,  according  to  Kohler,  1-155  »  from  vertical 
retorts,  according  to  Bueb,  i-ioo.  Most  coke-oven 
tars  have  specific  gravities  of  the  same  order,  i.e.,  lying 
between  i-i  and  1-2.  The  specific  gravity  is  influenced 
to  some  extent  by  the  amount  of  free  carbon  in  the  tar. 

(b)  Free  Carbon. — For  the  estimation  of  this  consti- 
tuent in  tar,  the  following  method,  due  to  Kraemer  and 
Spilker,  may  conveniently  be  adopted.  I  part  of  tar  is 
warmed  with  3  parts  of  aniline,  and  the  liquid  is  poured 
on  to  an  unglazed  porous  tile,  which  will  absorb  the 
soluble  part  of  the  tar  and  the  aniline,  leaving  the 
insoluble  carbon  as  a  flaky  mass.  The  latter  is  trans- 
ferred, without  loss,  to  a  weighed  watch  glass,  and 
weighed  after  drying  in  a  steam  oven  for  several  hours. 

As  was  pointed  out  above,  the  formation  of  free 
carbon  is  encouraged  by  high  distillation  temperatures  ; 
gas  tars  are  therefore  liable  to  contain  more  of  this 
constituent  than  coke-oven  tars.  The  carbon,  which  is 
deposited  as  a  hard  graphitic  mass,  remains  in  the  retort 
forming  part  of  the  coke  ;  part  of  the  finely  divided 
carbon,  which  is  formed  at  the  same  time,  passes  over 


38        Industrial  Organic  Analysis 

with  the  tar,  and  part  remains  in  the  pitch  after  the 
volatile  portions  have  been  distilled  off.  It  is  possible 
to  estimate  roughly  the  amount  of  pitch  which  may  be 
expected  from  a  tar  from  the  amount  of  free  carbon 
which  is  contained  in  the  latter,  as  good  pitch  of 
medium  hardness,  as  produced  by  most  tar  distillers, 
contains,  on  an  average,  about  28  per  cent,  of  free 
carbon.  Tars  containing  a  large  proportion  of  this 
constituent  are  apt  to  froth  on  distillation. 

(c)  Distillation  Test. — This  test  is  carried  out  with  a 
view  to  ascertaining  the  nature  and  approximate 
amounts  of  the  various  fractions  obtainable  on  distilling 
the  tar  on  a  large  scale  ;  as  it  is  hardly  possible  to 
reproduce  the  conditions  obtaining  on  a  manufacturing 
scale  in  an  ordinary  small  scale  laboratory  operation, 
it  is  customary  to  distil  from  3  to  5  kilos  of  tar  in  specially 
constructed  metal  vessels.  Such  methods  will  not  be 
described  here,  as  they  cannot  be  conveniently  carried 
out  with  the  ordinary  laboratory  equipment ;  for  a 
detailed  description,  including  also  the  estimation  of 
water  in  tar,  see  Lunge's  "  Technical  Methods  of 
Chemical  Analysis,"  Vol.  II.,  Part  II.,  pp.  761  et  seq. 

The  following  small  scale  operation,  devised  by 
B.  Nickels,  is  described,  as  it  will  afford  an  opportunity 
of  studying  the  behaviour  of  coal  tar  on  distillation,  and 
provide  further  material  for  analytical  work  :  250  c.c. 
of  tar  are  introduced  into  a  glass  retort  of  750  c.c. 
capacity  ;  the  retort  is  placed  on  a  cup-shaped  piece  of 
coarse  wire  gauze  which  rests  in  a  circular  hole  in  a  piece 
of  sheet  iron.  No  thermometer  or  special  condensing 
apparatus  are  necessary,  and  the  heat  is  supplied  by 
means  of  a  powerful  Bunsen  burner  which  is  protected 
from  draughts  by  asbestos  screens.  The  retort  is  covered 
by  a  dome,  which  may  be  made  from  a  tin  can,  a  trifle 
larger  than  the  bulb  of  the  retort,  by  cutting  out  a 
piece  from  its  side  in  order  to  make  room  for  the  neck. 
The  heating  is  regulated  so  that  the  distillate  falls  in 


Coal  Tar  and  its  Distillation  Products   39 

'drops  in  rapid  succession  ;  towards  the  end,  it  will  be 
necessary  to  heat  strongly,  so  that  the  wire  gauze 
becomes  red  hot  ;  when  the  pitch  begins  to  intumesce, 
the  heating  is  discontinued.  If  the  distillate  solidifies 
in  the  tubulure,  it  is  melted  down  by  cautiously  heating 
with  a  Bunsen  flame.  The  distillate  is  divided  into  the 
following  fractions  :  — 

(1)  Ammoniacal  liquor  and  total  light  oils. 

(2)  Middle  and  heavy  oils. 

(3)  Anthracene  oils. 

Fraction  (i)  is  collected  in  a  graduated  cylinder,  which 
is  changed  for  a  small  weighed  beaker  as  soon  as  a  drop 
of  the  distillate  solidifies  when  dropped  into  water.  The 
amount  of  total  light  oils  will  be  too  small  for  examina- 
tion ;  after  reading  off  their  volume,  they  are  separated 
from  the  ammoniacal  liquor,  which  may  then  be  titrated 
by  means  of  standard  acid,  in  order  to  estimate  the 
ammonia.  The  second  fraction  will,  at  first,  consist 
largely  of  solid  naphthalene,  and  will  afterwards  become 
more  liquid.  When  a  drop  of  the  distillate,  collected  on 
a  cold  metal  surface,  deposits  yellow  or  greenish  amor- 
phous matter,  the  receiver  is  changed  for  a  second  small 
weighed  beaker  in  which  the  last  fraction  is  collected 
until  the  heating  is  discontinued.  The  second  and  third 
fractions  may  be  weighed,  the  former  being  assayed  for 
phenols,  or  tar  acids,  and  the  latter  for  anthracene,  by 
methods  subsequently  described.  When  the  retort  is 
nearly  cold  it  is  plunged  into  cold  water  ;  the  pitch  will 
then  shrink,  so  that  it  may  be  removed  in  a  lump  on 
breaking  the  retort,  and  weighed.  Knowing  the  volume 
of  the  tar  distilled  and  its  specific  gravity,  the  per- 
centage amounts  of  the  various  constituents  weighed 
may  readily  be  calculated. 


4-O        Industrial  Organic  Analysis 

During  the  first  stages  of  the  distillation,  the  presence 
of  water  may  cause  bumping  ;  should  the  inconvenience 
caused  thereby  be  serious  the  tar  should  first  be  freed 
from  water  by  the  method  described  under  the  deter- 
mination of  the  specific  gravity. 

The  above  process  must,  of  course,  not  be  regarded  as 
an  exact  analytical  operation  ;  apart  from  the  fact  that 
unavoidable  losses  of  the  lighter  constituents  will  occur, 
it  must  be  borne  in  mind  that  the  relative  amounts  of 
the  various  fractions  may  vary  according  to  the  rate  of 
the  distillation.  Operations  of  this  nature  should 
always  be  carried  out  under  exactly  similar  conditions, 
if  comparable  results  are  desired. 


THE  EXAMINATION  OF  FIRST  RUNNINGS  AND  LIGHT  OILS. 

When  these  two  fractions  are  collected  together,  as  is 
often  the  case,  they  are  known  as  "  first  light  oils," 
"  crude  or  once  run  naphtha,"  or  "  total  light  oils." 
For  the  sake  of  clearness,  the  latter  term  will  be  employed 
to  signify  the  total  distillate,  exclusive  of  ammoniacal 
liquor,  resulting  from  the  first  distillation  of  coal  tar,  up 
to  the  point  at  which  the  distillate  becomes  heavier  than 
water.  The  products  coming  under  this  heading  are 
usually  yellow  to  dark  brown,  mobile  liquids  of  a  pene- 
trating smell  recalling  ammonium  sulphide,  carbolic  acid 
and  naphthalene,  at  the  same  time.  A  green  fluorescence 
is  sometimes  observable,  owing  to  tar  carried  over  in 
small  quantities  during  the  distillation.  The  total  light 
oils  are  usually  completely  volatile  below  180°  on 
redistillation,  and  have  specific  gravities  ranging  from 
0-910  to  0-950. 

According  to  Kraemer  and  Spilker  (Muspratt-Bunte's 
"  Chemistry,"  Vol.  VIII.,  p.  16),  the  composition  of  the 
total  light  oils  is  as  follows  : — 

Phenols,  consisting  of  phenol,  cresols  (chiefly  meta 
cresol)  and  small  quantities  of  xylenols,  5  to  15  per  cent. 


Coal  Tar  and  its  Distillation  Products   41 

Bases,  chiefly  pyridine  and  its  homologues,  i  to  3  per 
cent. 

Sulphur  compounds,  consisting  of  carbon  disulphide, 
thiophen  and  its  homologues,  o-i  per  cent. 

Nitriles,  such  as  aceto  and  benzo-nitriles,  0-2  to  0-3  per 
cent. 

Neutral  oxygen  compounds,  such  as  acetone  and 
coumarone,  i-o  to  1*5  per  cent. 

Hydrocarbons,  3  to  5  per  cent,  being  defines  from 
hexylene  and  upwards,  0-5  to  i-o  per  cent.,  paraffins 
beginning  from  hexane,  and  i-o  to  1-5  per  cent.,  unsatu- 
rated  compounds  which  combine  with  bromine  at  the 
ordinary  temperature,  such  as  cyclopentiadenes  and  the 
hydrobenzenes. 

The  remaining  80  per  cent,  consists  of  aromatic  hydro- 
carbons, of  which  about  four-fifths  belong  to  the  benzene 
series,  and  one-fifth  to  the  naphthalene  and  other  series, 
including  hydrocarbons  of  higher  molecular  weight. 
The  benzene  hydrocarbons  consist  approximately  of 
benzene,  100  parts  ;  toluene,  30  parts  ;  xylenes,  15  parts ; 
trimethyl  benzenes,  10  parts  ;  tetramethyl  benzenes, 
i  part ;  together  with  traces  of  higher  methylated  and 
ethylated  benzenes. 

Although,  as  will  be  gathered  from  the  above  list,  the 
light  oils  are  exceedingly  complex  mixtures,  the  tests  to 
which  they  are  usually  submitted  in  the  laboratory  are 
limited  to  (a)  the  determination  of  specific'  gravity, 
(b)  fractional  distillation,  and  (c)  the  estimation  of 
phenols.  These  tests  will,  at  any  rate,  enable  the  analyst 
to  distinguish  between  first  runnings,  light  oils  and  total 
light  oils  ;  in  order  to  fully  understand  the  significance 
of  the  results  obtained,  it  will,  however,  be  necessary  to 
have  considerable  experience  in  dealing  with  the  pro- 
ducts in  question,  both  in  the  laboratory  and  on  a  manu- 
facturing scale.  Further  chemical  tests  are  applied  to 
the  purer  products  obtained  from  the  first  runnings  and 
light  oils  on  distillation.  (See  below,  under  "  Benzols, 
Commercial  Benzene,  Toluene  and  Xylene.") 

(a)  Specific  Gravity. — This  may  be  determined  by 
means  of  an  ordinary  hydrometer  or  specific  gravity 


42        Industrial  Organic  Analysis 

float,  or  more  accurately,  by  means  of  a  Westphal  or 
Mohr  balance.  An  instrument  of  this  type  is  shown  in 
Fig.  4.  For  use,  it  is  mounted  as  shown,  the  thermometer 
plummet  being  suspended  in  distilled  water  at  15°  from 
the  end  of  the  graduated  beam  by  means  of  a  fine 
platinum  wire  ;  equilibrium  is  established  by  means  of 
the  adjusting  screw,  with  weights  corresponding  to  a 

specific  gravity  of 
i-oooo  on  the  beam. 
Each  of  the  largest 
weights  corresponds  to 
o-i  in  the  figure  for  the 
specific  gravity ;  one  of 
these  may  therefore  be 
suspended  from  divi- 
sion 10,  or  two  of  them 
from  division  5  when 
making  this  preliminary 
adjustment.  The  vessel 
containing  the  water  is 
then  removed,  and  the 


FIG.  4.— Westphal  Specific  Gravity 
Balance. 


plummet  is  wiped  dry 
and  suspended  in  the 
liquid  of  which  the 

specific  gravity  is  to  be  determined,  without  disturbing 
the  adjustment  of  the  instrument.  The  temperature  of 
the  liquid  should  be  the  same  as  that  of  the  water  which 
it  has  replaced.  The  pointer  of  the  balance  is  then 
brought  back  into  its  original  position  by  adjusting  the 
necessary  weights  on  the  graduated  arm  ;  as  each  weight 
weighs  ten  times  as  much  as  the  next  smaller  size,  the 
specific  gravity  may  be  directly  read  off  without  any 
calculation.  If  the  specific  gravity  float  is  used,  it  will 
also  be  necessary  to  have  the  liquid  at  15°. 


Coal  Tar  and  its  Distillation  Products   43 

The  specific  gravity  of  the  total  light  oils  usually  lies 
between  0-910  and  0-950.  The  average  specific  gravity 
of  English  first  runnings  is  0-900,  and  that  of  light  oils 
0-975.  The  indications  furnished  by  this  test  are  not  so 
definite  as  those  obtained  from 
the  two  following  tests. 

(b)  Distillation  Test.— For  the 
carrying  out  of  this  test,  several 
methods  are  in  use,  and  it  is, 
therefore,  necessary  to  exercise 
some  caution  in  comparing  re- 
sults obtained  by  different  ana- 
lysts. The  most  reliable  and 
scientifically  accurate  method  is 
that  generally  employed  in  Ger- 
many, which  consists  in  distilling 
the  liquid  from  a  flask  fitted 
with  an  effective  fractionating 
column.  When  such  an  appli- 
ance is  inserted  between  the 
distillation  flask  and  the  con- 
denser, the  higher  boiling  por- 
tions of  the  liquid  which  might 
otherwise  be  carried  over  with 
the  vapours,  are  condensed  and 
run  back  into  the  flask  ;  a  more 
complete  separation  of  the,  con- 
stituents according  to  their  boiling  points  is  thus 
obtained.  For  the  present  purpose,  Lunge  recommends 
the  use  of  a  three  bulb  Linnemann  apparatus  or  the 
Hempel  tube.  Sidney  Young's  "  Pear  "  and  "  Rod  and 
Disc  "  forms,  figured  here,  are  also  very  efficient  for 
the  fractionation  of  benzols. 

The  thermometer  should  be  placed  so  that  the  upper 
end  of  the  bulb  is  on  a  level  with  the  bottom  of  the 


FIGS.  5  and  6. — Sidney 
Young's  "  Rod  and 
Disc"  and  "  Pear" 
Fractional  i  ng 
Columns. 


44        Industrial  Organic  Analysis 

side  tube  of  the  column.  A  straight  tube  Liebig 
condenser  is  used,  and  the  distillate  is  received  in 
graduated  cylinders.  300  c.c.  of  the  liquid  may  be 
distilled  from  a  500  c.c.  flask,  and  the  volume  of  the 
distillate  noted  every  10°,  from  80°  and  onwards. 

From  the  results  thus  obtained,  an  approximate  idea 
may  be  formed  as  to  the  composition  of  the  liquid,  which 
will  give  some  indication  of  the  portions  of  the  coal  tar 
distillate  of  which  it  is  composed.  Working  with  larger 
quantities,  preferably  freed  from  carbon  disulphide,  as 
described  below,  benzene,  'b.p.  80°  to  81°,  and  toluene, 
b.p.  110°  to  in0,  may  be  separated  in  a  state  of  purity. 
Ortho,  meta  and  para-xylenes,  boiling  at  142°,  139°  and 
138°  respectively,  are  obtained  as  a  mixture ;  they 
cannot  be  separated  by  fractional  distillation. 

Good  first  runnings  should,  according  to  Lunge,  yield 
at  least  10  per  cent,  by  volume,  below  100°,  and  when 
the  product  collected  up  to  130°  is  redistilled,  it  should 
yield  at  least  25  per  cent,  of  its  volume  below  100°. 
First  runnings  yield,  on  an  average,  about  78  per  cent., 
by  volume,  below  171°. 

Light  oils  should  yield  little  below  120°,  and  about 
30  per  cent,  by  volume,  between  120°  and  171°.  All 
which  comes  over  above  the  latter  temperature  belongs, 
properly,  to  the  carbolic  oil  fraction.  If  an  appreciable 
amount  distils  below  120°,  the  oil  probably  contains  first 
runnings,  while  if  the  total  yield  up  to  171°  falls  con- 
siderably below  30  per  cent.,  a  portion  of  the  coal  tar 
distillate,  properly  belonging  to  the  carbolic  oil,  has 
probably  been  allowed  to  run  into  the  light  oil. 

(c)  Determination  of  Phenols. — The  following  rapid 
method  does  not  yield  results  of  great  accuracy,  but  may 
conveniently  be  adopted  in  technical  work.  It  is  based 
on  the  solubility  of  the  phenols,  and  the  insolubility  of 
the  neutral  oils  in  caustic  alkali  solution.  50  c.c.  of  the 
sample  are  introduced  into  a  graduated  glass-stoppered 


Coal  Tar  and  its  Distillation  Products    45 

cylinder  of  about  250  c.c.  capacity,  and  100  c.c.  of  a 
9  per  cent,  solution  of  sodium  hydroxide  are  gradually 
added.  The  whole  is  well  mixed  by  shaking,  and  allowed 
to  stand  until  it  has  separated  into  two  well  denned 
layers  ;  the  volume  of  the  neutral  oils  is  then  read  off 
and  subtracted  from  the  original  volume  of  the  sample  ; 
the  difference  is  taken  as  an  approximate  measure  of 
the  phenols  present.  A  more  accurate  reading  is  got 
by  adding  a  volume  of  petroleum  either  equal  to  that  of 
the  sample,  deducting  this  from  the  final  reading. 

More  accurate  methods  for  estimating  phenols  are 
given  below,  when  dealing  with  the  purer  products. 


Percentage  Distilling  below 

Sp.  Gr. 
at 
I5'5°  C. 

Approximate 
Composition. 

- 

100° 

c. 

120° 

C. 

f 

1  60° 
C. 

90    per    cent. 

90 

~ 

_ 

_ 

0-880 

70      per      cent. 

benzol. 

to 

benzene,  24  per 

0-888 

cent.       toluene, 

traces  of  xylene, 

4  —  6    per    cent. 

carbon       disul- 

phide   and  light 

paraffins,  etc. 

50    per    cent. 

5° 

90 

— 

— 

0-880 

Chiefly     toluene 

benzol. 

to 

and  xylene  with 

0-872 

a  little  benzene. 

30    per    cent. 

3° 

90 

— 

— 

0-875 

Chiefly     toluene 

benzol. 

and  xylene. 

Solvent  naphtha 

Nil 

— 

8—30 

90 

0-875 

Chiefly     xylene 

and      higher 

homologueswith 

a  little  naphtha- 

lene   and     par- 

affins, etc. 

Burning 

Nil 

— 

— 

30 

0-885 

Chiefly      xylene 

naphtha. 

and       higher 

homologues  ; 

naphthalene  and 

- 

paraffins,  etc. 

46        Industrial  Organic  Analysis 

Light  oil  is  used  as  such,  for  making  varnishes  for 
wood  and  iron,  and  occasionally  also  as  an  illuminant 
and  as  a  solvent  for  pitch.  By  far  the  most  important 
use  of  light  oil,  as  well  as  first  runnings,  is  in  the  manufac- 
ture of  the  "  benzols  "  of  commerce,  from  which  benzene, 
toluene  and  xylene  may  be  produced  in  a  state  of  purity 
for  the  manufacture  of  colours  and  other  valuable 
products. 

Before  redistillation,  the  first  runnings  or  light  oil  is 
washed,  first  with  dilute  sulphuric  acid  to  remove 
pyridine  bases,  then  with  concentrated  sulphuric  acid 
to  remove  defines  and  other  unsaturated  compounds 
such  as  coumarone,  indene,  cyclopentadiene,  etc.,  which 
are  converted  into  resins,  and  finally  with  dilute  alkali 
to  remove  phenols. 

The  chief  products  resulting  from  the  distillation  of 
the  washed  light  oils  of  tar  are  set  out  in  the  table  on 
p.  45,  together  with  their  behaviour  on  further  fractiona- 
tion,  specific  gravities  and  approximate  compositions. 
The  results  quoted  here  have  been  collected  from  data 
given  by  Lunge  and  Allen. 

THE    EXAMINATION    OF    BENZOLS    AND    COMMERCIAL 
BENZENE,  TOLUENE  AND  XYLENE. 

The  tests  commonly  applied  to  these  products,  some 
of  which  are  described  in  the  accompanying  table,  are 
(a)  the  determination  of  specific  gravity ;  (b)  the 
investigation  of  behaviour  on  fractional  distillation  ; 
(c)  the  estimation  of  carbon  disulphide  ;  and  (d)  the 
estimation  of  non-nitratable  hydrocarbons.  Carbon 
disulphide  occurs  chiefly  in  90  per  cent,  benzols  (see 
table),  and  only  in  comparatively  small  amounts  in  the 
higher  boiling  benzols ;  its  presence  in  appreciable 
quantities,  in  benzene  which  is  to  be  nitrated,  is  objec- 
tionable. Light  hydrocarbons,  including  defines,  con- 
sisting mainly  of  pentene,  and  paraffins,  occur  chiefly 
in  the  low-boiling  benzols,  though  higher  open  chain 
hydrocarbons  are  often  met  with  in  the  heavy  benzols 
and  in  commercial  toluene  and  xylene.  The  presence 
of  such  impurities  in  appreciable  amounts  -is  highly 


Coal  Tar  and  its  Distillation  Products    47 

objectionable,  causing  trouble  in  the  nitration  process, 
and  lowering  the  yield  of  amines.  Commercial  benzols 
are  sometimes  adulterated  with  petroleum  spirit,  which 
consists  chiefly  of  heptane,  or  with  shale  naphtha,  which 
contains  about  50  per  cent,  of  olefines,  mainly  heptene, 
and  about  50  per  cent,  of  paraffins. 

In  addition  to  the  above-mentioned  tests,  the  following 
tests  for  determining  the  relative  purity  of  benzene, 
toluene  and  xylene  are  described  :  (e)  the  bromine 
absorption  test,  and  (/)  the  sulphuric  acid  test.  Under 
the  heading  Xylene  will  be  described  the  estimation  of 
of  meta-xylene  in  presence  of  its  isomers. 

The  value  of  a  benzol  depends,  firstly,  on  its  relative 
freedom  from  the  undesirable  impurities  mentioned 
above,  and  secondly,  on  the  proportion  of  benzene  or 
toluene  present,  the  latter  substances  being  two  of  the 
most  valuable  constituents  of  coal  tar.  As  regards  the 
xylenes,  only  the  meta  isomer  is  of  any  value  to  the 
colour  manufacturer,  the  ortho  and  para  isomers  being 
looked  on  as  undesirable  impurities. 

(a)  Specific  Gravity. — This  is  best  determined  by 
means  of  the  Westphal  balance,  as  described  on  p.  42. 

The  indications  afforded  by  this  test  are  not  always  of 
a  very  definite  nature.  In  the  case  of  90  per  cent, 
benzol,  a  high  specific  gravity  may  generally  be  taken  as 
pointing  to  the  presence  of  an  appreciable  quantity  of 
carbon  disulphide.  The  specific  gravity  of  coal  tar 
naphtha  should  never  be  below  0-870  ;  if  lower,  light 
paraffins  are  probably  present,  as  the  specific  gravity 
of  petroleum  spirit  is,  at  the  most,  only  very  little  over 
0700.  On  the  other  hand,  the  effect  of  the  presence 
of  carbon  disulphide  on  the  specific  gravity  counter- 
acts that  of  the  light  paraffins,  so  that  a  sample  contain- 
ing both  of  these  impurities  may  well  have  a  normal 
specific  gravity. 


(* 

desc 


(b)  Fractional  Distillation. — This  may  be  carried  out  as 
ribed  under  the  heading  of  Light  Oils  and  First  Run- 
nings ;    the  volume  of  the  distillate  should  be  noted  at 


48        Industrial  Organic  Analysis 

the  temperatures  given  in  the  last  table,  for  comparison 
of  results. 

The  products  dealt  with  under  the  present  heading 
distil  within  narrower  limits  of  temperature  than  the 
crude  coal  tar  distillates  from  which  they  are  derived, 
the  constancy  of  boiling  point  of  a  sample  naturally 
depending  on  the  number  of  times  which  it  has  been 
redistilled  with  a  view  to  purification.  The  terms 
"90  per  cent,  benzol,"  "  50  per  cent,  benzol,"  etc.,  refer 
to  the  proportion  of  the  original  sample  which  distils 
below  and  up  to  100°.  According  to  Allen,  a  good 
90  per  cent,  benzol  should  not  distil  below  80°,  and  only 
yield  20  to  30  per  cent,  below  85°,  and  not  more  than 
90  per  cent,  below  100°.  It  should  be  wholly  distillable 
below  120°.  If,  say,  35  to  40  per  cent,  distils  below  85°, 
too  much  carbon  disulphide  or  light  hydrocarbons  are 
probably  present. 

50  per  cent,  benzol  should  distil  wholly  below  130°, 
and  should  yield  50  per  cent,  below  100°  and  40  per  cent, 
between  100°  and  120°. 

30  per  cent,  benzol  should  yield  30  per  cent,  below 
100°  and  60  per  cent,  from  100°  to  120° 

Pure  benzene,  toluene  and  xylene  of  commerce  distil 
within  i°  C. 

For  the  quantitive  estimation  of  the  pure  benzenoid 
hydrocarbons  on  a  large  scale,  see  Lunge's  "  Coal  Tar 
and  Ammonia,"  Vol.  II.,  pp.  774  et  seq. 

(c)  Carbon  Disulphide. — The  process  for  the  estima- 
tion of  this  constituent,  now  to  be  described,  depends  on 
the  formation  of  potassium  xanthate  by  the  interaction 
of  carbon  disulphide  and  potassium  ethoxide  as  follows: — 

CRl5/>0  +  CSa  =  S 

The  precipitated  xanthate  is  separated  and  estimated  by 
titration  with  standard  copper  sulphate  solution,  or 
analysed  for  sulphur. 

If  the  sample  should  be  turbid  owing  to  the  presence 


Coal  Tar  and  its  Distillation  Products   49 

of  water,  it  should  first  be  dehydrated  by  shaking  up 
with  plaster  of  Paris  and  filtering.  For  the  estimation, 
Kraemer  and  Spilker  recommend  the  following  process  : 
50  grams  of  benzol  are  mixed  with  50  grams  of  alcoholic 
potash,  made  by  dissolving  n  grams  of  potassium 
hydroxide  in  90  grams  of  absolute  alcohol ;  the  whole  is 
then  shaken  at  intervals  during  several  hours.  If  carbon 
disulphide  is  present,  the  xanthate  will  separate  out  in 
yellow  silky  needles.  The  latter  are  separated  in 
aqueous  solution  by  shaking  the  mixture  with  100  c.c.  of 
water  in  a  separating  funnel  and  washing  the  remaining 
benzol  with  successive  small  quantities  of  water  which 
are  added  to  the  main  aqueous  extract. 

The  xanthate  may.  then  be  estimated  in  the  aqueous 
solution  as  follows  :  A  solution  containing  12-475  grams 
of  crystallised  copper  sulphate  per  litre  is  prepared ; 
i  c.c.  of  this  corresponds  to  0-0076  gram  of  carbon 
disulphide  in  the  titration  described  below.  The  aqueous 
xanthate  solution  is  acidified  with  dilute  acetic  acid, 
whereupon  it  must  immediately  be  titrated  with  the 
copper  solution,  as  free  xanthic  acid  decomposes  spon- 
taneously. A  brown  precipitate  of  cuprous  xanthate 
will  be  formed  at  first ;  on  the  addition  of  more  of  the 
copper  solution,  this  precipitate  is  transformed  into  the 
bright  yellow  cupric  xanthate.  The  end  point  is  reached 
when  a  drop  of  the  solution  taken  out  on  a  glass  rod,  and 
placed  beside  a  drop  of  dilute  potassium  ferrocyanide 
solution  on  a  piece  of  filter  paper,  produces  a  red  mark 
at  the  point  of  contact  of  the  liquids ;  the  amount  of 
copper  solution  which  has  been  added  should  then  be  in 
slight  excess  of  that  required  to  interact  with  the  whole 
of  the  xanthic  acid  present,  to  form  the  insoluble  cupric 
xanthate.  The  amount  of  carbon  disulphide  in  the 
sample  may  then  be  calculated  from  the  number  of  c.c. 

I.O.A.  4 


50        Industrial  Organic  Analysis 

of  copper  solution  used,  by  employing  the  factor  given 
above. 

The  following  alternative  methods  may  also  be 
employed : — 

(i.)  The  acidified  solution  of  potassium  xanthate  is 
treated  with  an  excess  of  copper  sulphate  solution,  and 
the  precipitated  cupric  xanthate  filtered  off,  washed, 
ignited  and  determined  as  cupric  oxide  or  copper,  as 
usual. 

(ii.)  The  potassium  xanthate  solution  is  warmed  with 
an  excess  of  potassium  hydroxide  solution  and  bromine 
till  perfectly  clear ;  the  sulphur,  which  will  now  all  be 
present  as  alkali  sulphate,  is  determined  by  precipitation 
with  barium  chloride  in  presence  of  hydrochloric  acid  in 
the  usual  way. 

After  the  removal  of  the  carbon  disulphide,  the 
residual  benzol  should  be  examined  by  distillation  and 
tested  for  its  specific  gravity  ;  a  reduction  both  in  the 
amount  of  distillate  coming  over  below  85°  and  in  the 
specific  gravity  should  be  noticeable  if  the  sample 
contained  an  appreciable  amount  of  the  impurity. 

The  quantity  of  alcoholic  potash  recommended  above 
is  sufficient  for  the  removal  of  quantities  up  to  5  per  cent, 
of  carbon  disulphide.  In  very  exceptional  cases  only 
will  the  percentage  of  carbon  disulphide  in  90  per  cent, 
benzol  exceed  this  limit ;  usually  it  only  amounts  to 
I  to  2,  per  cent.  In  50  per  cent,  benzol  it  may  sometimes 
amount  to  as  much  as  I  per  cent.,  while  in  the  higher 
boiling  products,  it  is  either  entirely  absent  or  present 
in  traces.  The  pure  benzene  of  commerce  may  contain 
from  about  o-i  to  0-3  per  cent,  of  carbon  disulphide.  As 
was  mentioned  above,  benzene  or  toluene  which  is  to  be 
nitrated  should  be  as  free  as  possible  from  this  impurity. 

(d)  Non-Nitratable  Hydrocarbons.  —  The  so-called 
"  nitrofication  test,"  for  the  estimation  of  benzene,  and 
non-nitratable  hydrocarbons,  by  nitrating  the  benzol, 


Coal  Tar  and  its  Distillation  Products    51 

will  not  be  described  here,  owing  to  its  inaccuracy.  The 
following  method,  due  to  Frank,  Kraemer  and  Spilker, 
is  based  on  the  fact  that  the  benzenoid  hydrocarbons  are 
converted,  at  the  ordinary  temperature,  into  water1 
soluble  sulphonic  acids  ;  the  oily  residue  which  is  unacted 
on  by  the  acid  may  consist  of  paraffins,  naphthenes 
(i.e.,  hydrogenated  aromatic  hydrocarbons)  and  carbon 
disulphide. 

200  grams  of  the  sample  are  placed  in  a  separating 
funnel,  and  500  grams  of  fuming  sulphuric  acid,  contain- 
ing 20  per  cent,  of  the  anhydride,  are  cautiously  added 
in  small  amounts,  shaking  well  after  each  addition. 
When  the  whole  of  the  acid  has  been  added,  the  mixture 
is  shaken  for  15  minutes,  and  then  allowed  to  stand  for 
2  hours  to  separate.  The  lower  acid  layer  is  run  off,  and 
the  upper  layer  is  treated  with  two  successive  portions 
of  500  grams  of  the  fuming  acid  as  just  described.  The 
oily  layer  is  separated  off,  and  the  acid  extracts  are 
united  and  run  slowly  and  cautiously  on  to  an  equal 
weight  of  pourided  ice.  Care  should  be  taken  that  the 
temperature  does  not  rise  above  40°.  The  diluted  acid 
will  contain  a  certain  quantity  of  oily  residue,  unacted 
on  by  the  fuming  acid,  which  has  either  been  dissolved  by 
the  sulphonic  acids,  or  mechanically  removed  from  the 
main  portion.  This  is  separated  by  distilling  the 
mixture  from  a  flask,  over  a  free  flame,  the  distillate 
being  caught  in  a  small  separating  funnel.  When,  apart 
from  any  oil  which  passes  over,  50  c.c.  of  liquid  have  been 
collected,  the  distillation  is  discontinued,  and  the  oily 
portion  of  the  distillate  separated  from  the  aqueous 
layer  and  added  to  the  main  portion  of  the  oily  matter 
which  was  unacted  on  by  the  fuming  acid.  The  latter  is 
treated  with  successive  portions  of  30  c.c.  of  fuming 
sulphuric  acid  as  described  above,  until  no  further 
sensible  diminution  in  volume  takes  place.  The  residual 

4—2 


52        Industrial  Organic  Analysis 

oil  is  washed  with  a  small  quantity  of  distilled  water  in 
the  funnel,  separated  carefully,  transferred  to  a  tared 
flask  and  weighed.  If  the  carbon  disulphide  has 
previously  been  determined,  it  may  be  deducted  from 
the  total  non-nitratable  matter.  The  remainder  will 
consist  of  paraffins,  and  possibly  also  of  naphthenes. 

By  this  method,  the  presence  of  an  undesirable  amount 
of  paraffins,  which  in  some  cases  may  be  due  to  adultera- 
tion with  petroleum  spirit  or  shale  oil,  may  be  detected. 
Benzols  should  only  contain  a  few  tenths  per  cent,  of 
paraffins,  or  at  the  most,  only  I  per  cent.  Commercial 
xylenes,  however,  sometimes  contain  up  to  3  per  cent, 
of  this  impurity. 

The  following  tests  are  sometimes  applied  in  order  to 
determine  the  relative  purity  of  commercial  benzene  and 
toluene  or  benzols  : — 

(e)  Bromine  Absorption  Test. — This  test,  devised  by 
Frank,  Kraemer  and  Spilker,  gives  an  indication  of  the 
amount  of  hydrocarbons  present,  which  combine  with 
bromine  at  the  ordinary  temperature,  i.e.,  unsaturated 
compounds,  which  should  have  been  removed  during 
the  treatment  of  the  benzol  with  strong  sulphuric  acid. 
(See  p.  46.)  It  is  not  applicable  to  xylene  or  to  mix- 
tures containing  much  of  this  constituent. 

A  tenth  normal  bromine  solution  is  prepared  by 
dissolving  9-9167  grams  of  potassium  bromide  and 
27833  grams  of  potassium  bromate  in  water  and  making 
up  to  i  litre ;  I  c.c.  of  this  solution  will  liberate  0-008 
gram  of  bromine  on  acidification  with  dilute  sulphuric 
acid.1 

5  c.c.  of  the  sample  are  placed  in  a  stoppered  bottle 
of  about  50  c.c.  capacity,  together  with  10  c.c.  of  20  per 
cent,  sulphuric  acid,  and  as  much  of  the  bromide  and 
bromate  solution  as  the  sample  will  decolorise  after 
shaking  uninterruptedly  for  5  minutes  is  quickly  added. 

1  5  KBr  +  KBrO3  +  6  HaSO*  -  6  KHSO4  +  3  H2O  +  3  Bra. 


Coal  Tar  and  its  Distillation  Products    53 

The  end  point  is  indicated  when  the  oil  floating  on  the 
top  shows  an  orange  red  after  standing  for  15  minutes, 
and  a  drop  of  it  momentarily  produces  a  dark  blue 
colour  on  zinc  iodide  and  starch  paper.  Preliminary 
trials  must  be  made  in  order  to  determine  how  much  of 
the  bromine  solution  will  be  required,  for  in  the  actual 
determination  it  is  necessary  that  the  full  amount  should 
be  added  at  once,  as  directed  above. 

Pure  benzene  or  toluene  of  commerce  should  give  a 
distinct  permanent  colour  after  adding  only  o-i  c.c.  of 
the  bromine  solution.  Commercial  50  per  cent,  or 
90  per  cent,  benzols  require,  on  an  average,  0-6  c.c.  and 
rarely  more  than  i  c.c. 

(/)  Sulphuric  Acid  Test. — This  test  is  also  due  to 
Frank,  Kraemer  and  Spilker ;  it  gives  an  indication  of 
the  relative  amount  of  matter  present  which  will  react 
with  strong  sulphuric  acid  at  the  ordinary  temperature. 
(Compare  introductory  remarks  to  the  bromine  absorp- 
tion test.) 

5  c.c.  of  the  sample  are  vigorously  shaken  with  5  c.c. 
of  concentrated  sulphuric  acid  in  a  stoppered  bottle,  for 
5  minutes,  and  then  compared  with  a  solution  of 
potassium  dichromate  in  50  per  cent,  sulphuric  acid, 
contained  in  a  similar  bottle.  50  and  90  per  cent, 
benzols  should  exhibit  a  colour  like  that  of  a  solution 
containing  0-5  to  at  most,  1-5  gram  of  chemically  pure 
potassium  dichromate  per  litre.  Xylene  will  give  a 
colour  like  a  solution  containing  1*2  to  2'O  grams  per 
litre,  while  pure  benzene  or  toluene  of  commerce  should 
give  no  colour  at  all. 

XYLENE. 

The  xylene  as  obtained  by  the  distillation  of  coal  tar 
contains  the  three  isomeric  dimethyl  benzenes  in  varying 


54        Industrial  Organic  Analysis 

roportions,  together  with  smaller  quantities  of  the 
igher  benzene  homologues  and  paraffins.  Of  these 
constituents,  meta  xylene  is  the  only  one  of  any  use  in 
the  manufacture  of  dyestuffs ;  the  ortho  and  para 
isomers  are  not  only  useless  for  this  purpose,  but  have 
to  be  removed  before  the  chemical  treatment  of  the  meta 
xylene  is  proceeded  with.  Paraffins  are  a  very  undesir- 
able impurity  and  may  render  the  xylene  unfit  for 
colour  making,  if  present  in  large  amounts.  The  follow- 
ing variations  in  composition  of  xylenes  from  English 
and  Scotch  tars  were  found  by  Levinstein  : — Paraffins, 
3  to  10  per  cent.,  ortho  xylene,  3  to  15  per  cent.,  para 
xylene,  3  to  10  per  cent.,  and  meta  xylene,  70  to  87  per 
cent. 

In  practice,  most  of  the  ortho  xylene  is  usually  removed 
as  sulphonic  acid  in  the  treatment  of  the  benzol  with 
sulphuric  acid  (see  p.  46) ;  meta  xylene  may  be  separated 
from  its  isomers  by  converting  the  mixed  xylenes  into 
sulphonic  acids  and  then  steam  distilling,  when  only 
the  meta  xylene  will  be  regenerated. 

Determination  of  Meta  Xylene  and  Paraffins  in  Crude 
Xylene. — The  method  described  here  is  due  to  Crafts. 
A  weighed  quantity  of  xylene,  about  10  to  20  grams,  is 
poured  on  to  two  and  a  half  times  its  weight  of  concen- 
trated sulphuric  acid  contained  in  a  tube  of  hard  glass  ; 
the  depth  of  the  xylene  layer  in  millimetres  is  noted, 
after  which  the  tube  is  sealed  and  heated  to  120°  for 
I  hour,  the  contents  being  well  mixed  by  shaking  from 
time  to  time.  After  cooling,  the  tube  is  opened,  and  the 
contents  are  treated  with  3  to  4  times  their  bulk  of  a 
mixture  of  equal  parts  of  concentrated  hydrochloric 
acid  and  water,  shaken  well  and  allowed  to  stand  for 
i  hour  at  the  ordinary  temperature.  The  insoluble  oily 
layer,  which  consists  of  saturated  hydrocarbons,  chiefly 
paraffins,  is  measured  in  the  tube  by  noting  its  depth  in 
millimetres,  or  better  still,  separated  off  in  a  funnel, 
distilled  and  weighed.  The  solvent  acid  is  returned  to 


Coal  Tar  and  its  Distillation  Products    55 

the  tube,  which  is  then  resealed  and  heated  to  122° 
for  20  hours.  After  this  treatment,  an  oily  layer  will 
have  been  formed  which  will  consist  of  approximately 
97  per  cent,  of  the  meta  xylene  ;  this  may  be  measured, 
or  separated,  distilled  and  weighed ;  of  the  xylene 
sulphonic  acids  formed  during  the  heating  with  sulphuric 
acid,  only  that  derived  from  the  meta  xylene  is  decom- 
posed by  heating  with  hydrochloric  acid  under  the 
conditions  of  the  experiment.  For  the  estimation  of 
ortho  and  para  xylenes  and  ethyl  benzene  by  a  continua- 
tion of  this  process,  see  Crafts,  Comptes  Rendus, 
114,  p.  mo. 

DISTINCTION  BETWEEN  COAL  TAR  BENZOL,  PETROLEUM 
SPIRIT  AND   SHALE  NAPHTHA. 

These  products,  though  similar  as  regards  appearance 
and  boiling  points,  may  be  distinguished  by  their  smell 
and  specific  gravities.  Thus  the  specific  gravity  at 
15°  of  benzols  is  practically  always  above  0-85,  while 
the  specific  gravities  of  petroleum  spirit  and  shale  oil  are 
considerably  lower,  usually  about  0700  and  072,  at 
15°  respectively.  The  chemical  distinction  already 
alluded  to  is  based  on  the  fact  that  the  paraffins  are 
unacted  on  by  fuming  sulphuric  acid  at  the  ordinary 
temperature,  while  the  aromatic  hydrocarbons  are 
converted  into  sulphonic  acids.  The  resistance  of  the 
former  to  the  nitrating  action  of  a  mixture  of  nitric  and 
sulphuric  acids  may  also  be  mentioned  here.  Petroleum 
spirit  contains  over  75  per  cent,  of  paraffins,  the  rest 
being  mainly  olefines ;  shale  naphtha  consists  of 
approximately  equal  parts  of  paraffins  and  olefines. 
The  latter  are,  as  is  well  known,  absorbed  by  strong 
sulphuric  acid,  usually  with  some  darkening  in  colour. 
The  following  simple  tests  may  be  used  for  the 


56        Industrial  Organic  Analysis 

differentiation  of  benzol  on  the  one  hand  and  petroleum 
spirit  and  shale  naphtha  on  the  other  :  (i.)  Coal  tar 
pitch  dissolves  freely  in  the  former,  but  only  slightly  in 
the  latter,  producing  a  comparatively  feeble  colouration  ; 
the  distinction  may  be  made  still  sharper  by  using  pitch 
which  has  previously  been  exhausted  with  petroleum 
spirit,  (ii.)  Picric  acid  dissolves  in  the  aromatic  hydro- 
carbons giving  a  yellow  solution,  but  is  insoluble  in 
the  two  other  products  under  consideration.  (See  also 
Chapter  V.,  p.  163.) 

THE  EXAMINATION  OF  MIDDLE  OR  CARBOLIC  OIL  AND 
ITS  PRODUCTS. 

The  principal  constituents  of  this  fraction  are  phenols 
and  naphthalene,  of  which  phenol  itself,  i.e.,  mono- 
hydroxy  benzene,  is  the  most  valuable.  At  the  ordinary 
temperature  a  large  portion  of  the  naphthalene  crystal- 
lises out  from  the  oil ;  at  40°,  middle  oil  is  a  brownish 
yellow  liquid,  smelling  of  carbolic  acid  and  naphthalene. 
Its  specific  gravity  at  15°,  which  may  be  determined  as 
described  for  crude  coal  tar,  generally  lies  between 
i -oo  and  1-03  ;  if  under  the  lower  limit  an  unduly  large 
amount  of  light  oils  may  be  present.  The  crude  naphtha- 
lene present  amounts  to  30  per  cent,  or  more.  In 
practice  the  greater  part  of  the  naphthalene  is  separated 
by  filtration.  The  remaining  carbolic  oil,  which  contains 
phenol,  cresols,  xylenols  and  other  higher  phenols, 
neutral  tar  oils,  water,  naphthalene,  pyridine  bases,  etc., 
is  either  sold  as  such  for  disinfecting  purposes  or  varnish 
making,  or  it  is  treated  with  caustic  alkali  solution  in 
order  to  separate  the  phenols  from  the  neutral  and  basic 
constituents.  The  phenols  are  recovered  from  the 
alkaline  solution  by  precipitation  with  mineral  acid,  and 
worked  up  for  carbolic  or  cresylic  acids  of  varying 
degrees  of  purity,  by  fractional  distillation  and  other 
processes  ;  the  portion  insoluble  in  alkali  is  worked  up 
for  heavy  solvent  naphtha,  pyridine  bases,  and  naphtha- 
lene. 


Coal  Tar  and  its  Distillation  Products    57 

Crude  Naphthalene  in  Middle  Oil. — 500  grams  of  the 
middle  oil  are  cooled  and  the  naphthalene  is  separated 
by  nitration,  pressed  between  filter  paper  until  no  longer 
oily,  and  weighed.  It  should  distil  mainly  between 


CRUDE  CARBOLIC  ACID. 

This  includes  the  portion  of  the  middle  oil  which  is 
liquid  at  the  ordinary  or  lower  temperatures,  having  been 
separated  from  the  solid  crude  naphthalene  by  nitration. 
According  to  Lunge,  it  should  have  an  average  boiling 
point  of  250°,  a  specific  gravity  between  0-99  and  i-oi, 
and  contain  from  25  to  35  per  cent,  of  phenols  and  about 
5  per  cent,  of  pyridine  bases.  It  may  be  examined  for 
(a)  total  phenols  and  water,  and  (b)  phenol,  i.e.,  the 
monohydroxy  benzene. 

(a)  Total  Phenols  and  Water. — The  following  approxi- 
mate method  is  due  to  Bach  ;  advantage  is  taken  of  the 
insolubility  of  the  phenols  in  brine,  and  their  solubility 
in  sodium  hydroxide  solution. 

50  c.c.  of  the  sample  are  distilled  from  a  retort  until 
solid  matter  begins  to  come  over,  the  distillate  being 
caught  in  a  clean  wide  100  c.c.  burette,  graduated  in 
fifths  of  a  c.c.  and  furnished  with  a  glass  tap.  Before  the 
distillation  25  c.c.  of  a  saturated  solution  of  common  salt 
are  introduced  into  the  burette,  and  the  level  of  the 
liquid  is  noted.  If  the  carbolic  acid  contains  no  water 
the  oil  separates  clearly  from  the  brine,  while  if  the 
contrary  is  the  case,  an  emulsion  is  formed  which  may, 
however,  be  broken  down  by  gently  agitating  the  liquid 
and  allowing  to  stand.  When  the  layers  have  become 
distinct,  the  level  of  the  brine  is  read  off ;  its  increase  in 
volume  is  a  measure  of  the  water  in  the  sample.  The 
brine  is  then  drawn  off,  the  level  of  the  oil  noted  and  the 
burette  filled  to  the  zero  mark  with  sodium  hydroxide 


58        Industrial  Organic  Analysis 

solution  of  specific  gravity  1-26  ;  after  closing  with  a 
cork  the  contents  are  mixed  by  shaking  vigorously  and 
then  allowed  to  settle.  If  the  burette  was  originally 
clean  and  free  from  grease  the  oil  will  have  separated  after 
half  an  hour,  when  its  level  may  be  observed.  The 
difference  between  this  and  the  previously  observed 
volume  of  the  oil  gives  the  amount  of  phenols  in  the 
sample. 

The  percentage  of  phenols  in  crude  carbolic  acid  is 
very  variable,  especially  as  the  inferior  qualities  are  often 
adulterated  with  neutral  tar  oils.  The  percentage  of 
phenols  is  usually  stated  in  naming  the  product,  e.g., 
"  carbolic  acid,  25  per  cent./'  "  carbolic  acid,  50  per 
cent." 

There  is  no  satisfactory  method  of  determining  the 
naphthalene  or  the  relative  proportions  of  phenol, 
cresols  and  higher  phenols  in  crude  carbolic  acid,  though 
the  following  process,  due  to  Chas.  Lowe,  will  furnish 
results  which  will  give  a  rough  idea  of  the  proportion  of 
phenol  present  in  a  sample  which  has  not  been  adulterated 
with  neutral  oils. 

(b)  Phenol. — This  method  combines  separation  by 
distillation  with  the  determination  of  solidifying  point  of 
the  distillate.  Comparison  is  made  with  the  solidifying 
points  of  mixtures  containing  known  proportions  of  pure 
phenol  and  cresols  ;  for  this  purpose  Cal vert's  pure 
products  may  be  used. 

100  c.c.  of  the  sample  are  distilled  from  a  retort,  and 
the  distillate  is  collected  in  graduated  tubes.  At  first 
water  distils  and  then  an  oily  fluid  ;  when  10  c.c.  of  the 
latter  have  been  collected,  the  receiver  is  changed.  The 
volume  of  the  water  is  read  off ;  if  the  oily  liquid  floats 
on  the  water,  it  is  light  oil  of  tar  ;  if  it  sinks  it  may  be 
regarded  as  hydrated  phenol,  containing  about  50  per 
cent,  of  phenol.  The  next  portion  which  distils  consists 


Coal  Tar  and  its  Distillation  Products    59 

of  anhydrous  phenols  ;  when  it  measures  62-5  per  cent., 
the  receiver  is  again  changed.  The  residue  is  wholly 
cresols  and  higher  phenols.  The  fraction  consisting  of 
62-5  per  cent,  is  cooled  and  its  solidifying  point  deter- 
mined ;  this  should  lie  between  15-5°  and  24°.  The 
proportion  of  phenol  to  cresols  in  the  fraction  may  then 
be  estimated  by  determining  the  solidifying  points  of 
synthetic  mixtures  of  pure  phenol  and  coal  tar  cresol. 


Crude  Carbolic  Acid  from  Tar 

from 

Black- 

Manches- 

Manches- 

burn. 

ter. 

ter. 

Water,  per  cent  .  ,  by  volume 
First  oil  up  to  185  (to  be 

12 

13 

15 

rejected)  per  cent. 

II 

II 

IO 

Carbolic     acid     distilling 

below  190°  per  cent.    . 

48 

45 

45 

Ditto  above  190° 

I3i 

17* 

17! 

Solidifying   points   of   the 

latter,  i.e.,  61  to  62  per 

cent,  of  the  total  sample. 

15° 

18° 

i6|° 

The  solidifying  point  may  be  determined  as  described 
under  "  Crystallised  Carbolic  Acid."     (See  p.  63.) 

The  proportion  of  light  oils  in  crude  carbolic  acid  is 
usually  small,  at  the  most  5  to  6  per  cent.,  hence  an 
appreciable  quantity  of  phenol  may  be  lost  in  the  first 
10  per  cent,  of  oily  matter  distilled.  In  some  cases, 
therefore,  it  may  be  best  to  take  the  solidifying  point  of 
the  distillate  coming  over  between  185  and  195°.  When 
dealing  with  better  qualities,  i.e.,  those  containing  more 
phenol,  it  is  recommended  to  take  the  solidifying  point 


60        Industrial  Organic  Analysis 

of  the  distillate  coming  over  between  180  and  190°,  as  in 
such  cases  the  whole  of  the  phenol  will  naturally  come 
over  at  lower  temperatures.  The  temperatures  referred 
to  are  indicated  by  a  thermometer,  the  bulb  of  which  is 
placed  in  the  vapour  of  the  boiling  liquid. 

The  table  on  p.  59  contains  results  obtained  by 
Watson  Smith  by  the  method  just  described. 

The  neutral  oils  may  amount  to  as  much  as  10  per  cent., 
and  the  water  may  lie  between  10  and  17  per  cent.  If 
an  excess  of  cresylic  acid  is  present,  crystallisation  is 
prevented  in  the  determination  of  the  solidifying  point ; 
in  this  case,  a  second  fractional  distillation  should  be 
made,  this  time  stopping  when  the  thermometer,  placed 
in  the  vapour,  indicates  190°. 

The  melting  and  boiling  points  of  phenol  and  the 
three  cresols  are  as  follows  : — 


— 

Phenol. 

Ortho 
Cresol. 

Meta  Cresol. 

Para 
Cresol. 

Melting  points  °  C.     . 
Boiling  points  °  C 

42-5—43 
182 

30—31 
IQO'5 

4 
200  —  20  I 

36-5 
2OI'8 

The  relative  proportions  of  the  three  cresols  in  coal 
tar  are  about  35  per  cent,  ortho,  40  per  cent,  meta  and 
25  per  cent.  para.  The  mixture  of  cresols  from  coal  tar 
boils  from  198°  to  203°. 

PUR£    CARBOLIC    ACID    AND    ITS    PREPARATIONS. 

The  crystallised  pure  carbolic  acids  of  commerce 
consist  of  more  or  less  pure  phenol ;  they  may  contain 
small  quantities  of  cresols  and  higher  boiling  compounds 
which  produce  a  red  or  yellow  colour,  water  and  traces 
of  metallic  compounds. 

Liquefied  pure  carbolic  acid  usually  consists  of  about 
o  parts  of  pure  phenol  and  10  parts  of  water  or  alcohol, 
t  may  be  distinguished  from  "  liquid  carbolic  acid," 


Coal  Tar  and  its  Distillation  Products    61 

which  usually  consists  of  cresols  and  higher  homologues 
by  the  two  following  tests  (a)  and  (b)  :— 

(a)  Boiling  point. — Liquefied  pure  carbolic  acid  begins 
to  boil  below  or  near  100°,  after  which  the  boiling  point 
quickly  rises  to  185°  to  190°,  while  the  product  containing 
cresols  will  boil  at  185°  to  209°. 

(b)  Solubility  in  water. — The  liquefied  product  requires 
at  most  18  parts  of  water  to  give  a  clear  solution,  while 
cresylic  acid  is  not  completely  dissolved  by  even  50  parts 
of  water. 

Estimation  of  Phenol. — The  method  to  be  described  is 
Koppeschaar's  modification  of  Landolt's  process,  based 
on  the  formation  of  the  insoluble  tribromophenol  by  the 
interaction  of  phenol  and  bromine  in  aqueous  solution. 
A  known  amount  of  bromine  having  been  added  to  the 
phenol  solution,  potassium  iodide  solution  is  added,where- 
upon  an  amount  of  iodine  corresponding  to  the  excess  of 
free  bromine  present  is  liberated ;  this  is  estimated  by 
titration  with  standard  sodium  thiosulphate  solution, 
using  starch  as  indicator.  The  amount  of  bromine  used 
up  in  the  formation  of  tribromophenol  may  then  be 
calculated.  It  should  be  noted  that  this  method  is  only 
applicable  to  the  products  dealt  with  under  the  present 
heading,  i.e.,  pure  or  nearly  pure  phenol,  or  solutions  of 
the  latter  which  do  not  contain  cresols  or  other  substances 
which  also  react  with  bromine  water.  The  following 
solutions  will  be  required  : — 

A  solution  of  sodium  thiosulphate,  corresponding  to 
5  grams  of  iodine  per  litre  (5  grams  I  =  9764  grams 
Na2S2O3  5  H2O).  This  solution  may  be  standardised 
according  to  Volhardt's  method,  as  described  on  p.  92. 

Potassium  iodide  solution. — A  10  per  cent,  solution  of 
the  pure  salt  in  water. 

Starch  solution. — Freshly  prepared  by  heating  half 


62        Industrial  Organic  Analysis 

a  gram  of  powdered  starch  with  50  c.c.  of  water  in  boiling 
water. 

Bromine  solution.  — A  solution  containing  5  molecular 
proportions  of  potassium  bromide  to  one  of  potassium 
bromate,  of  such  a  strength  that  50  c.c.  mixed  with  5  c.c. 
of  strong  hydrochloric  acid  and  100  c.c.  of  water1  requires 
for  complete  decolorisation  86  to  95  c.c.  of  the  thio- 
sulphate  solution  described  above.  It  may  be  prepared 
either  by  dissolving  the  pure  salts  in  water,  in  the 
requisite  proportions,,  or  by  adding  to  a  solution  of  pure 
sodium  hydroxide  an  excess  of  bromine  and  evaporating 
to  dryness ;  9  grams  of  the  powdered  residue  are  dis- 
solved in  100  c.c.  of  water  and  diluted  to  the  requisite 
strength  after  a  preliminary  titration. 

The  actual  determination  of  phenol  is  carried  out  as 
follows  :  4  grams  of  the  sample,  or  more  if  the  amount 
of  phenol  present  is  small,  is  dissolved  in  or  mixed  with 
water,  and  made  up  to  I  litre.  25  c.c.  of  this  solution, 
filtered  if  necessary,  are  placed  in  a  bottle  provided  with  a 
well  fitting  stopper,  of  about  400  c.c.  capacity.  100  c.c. 
of  the  bromide  and  bromate  solution  are  added,  and  then 
5  c.c.  of  concentrated  hydrochloric  acid,  in  order  to  set 
free  the  bromine.  The  bottle  is  immediately  closed, 
shaken  and  allowed  to  stand  for  15  minutes.  10  c.c. 
of  the  potassium  iodide  solution  are  added  and  the 
whole  is  well  mixed  again.  The  free  iodine  liberated 
by  the  excess  of  bromine  is  estimated  by  titration 
with  the  thiosulphate  solution,  adding  a  few  c.c.  of 
the  starch  solution  as  indicator  towards  the  end  of  the 
process. 

If  the  standard  solutions  used  are  of  the  strength 
prescribed  above,  and  the  operations  carried  out  as 

1  See  footnote,  p.  52.     2  KI  +  Br2  =  2  KBr  +  la. 

2  NaaSaOs  +  la  =  NaaS^Oe  a  Nal. 


Coal  Tar  and  its  Distillation  Products    63 

described,  then  the  percentage  of  phenol  in  the  sample  is 
given  by  the  formula 

(2  a— b)  X  0-61753, 

where  a  =  the  number  of  cubic  centimetres  of  thio- 
sulphate  solution  required  by  50  cc.  of  the  bromide 
and  bromate  solution  used,  and  b  the  number  of  cubic 
centimetres  required  by  the  iodine  equivalent  to  the 
final  excess  of  bromine. 

Care  should  be  taken  that  the  25  c.c.  of  solution  used 
for  titration  do  not  contain  more  than  o-i  gram  of 
phenol. 

Phenol  in  mixtures  with  Cresols, — For  the  approximate 
estimation  of  phenol  in  mixtures  with  cresols,  the  method 
of  determining  the  solidifying  point  of  the  fraction 
distilling  between  180°  and  190°  may  be  employed. 
(See  under  "  Crude  Carbolic  Acid,"  p.  59.) 

A  modification  of  the  above  titration  method  is 
described  in  Chapter  IV.  (p.  148)  for  the  determination 
of  phenol  and  cresols  in  phenolic  soaps. 

CRYSTALLISED   CARBOLIC   ACID.    TESTS   FOR   PURITY. 

Under  this  heading  are  described  the  determination 
of  (a)  the  solidifying  point,  and  (b)  water. 

(a)  Solidifying  Point. — This  is  the  most  important 
test  for  the  purity  of  crystallised  phenol.  The  purest 
phenol  solidifies  at  40-9°,  while  for  pure  phenol  of 
commerce,  it  is  usual  to  demand  a  solidifying  point  of 
39°  to  41°.  The  presence  of  cresols  or  water  causes  a 
lowering  in  the  solidifying  point.  For  the  determination, 
Kraemer  and  Spilker  give  the  following  directions  :  an 
average  sample  of  50  grams,  which  has  been  taken  from 
a  large  sample  which  has  been  liquefied  and  mixed  with 
precautions  for  the  exclusion  of  moisture,  is  allowed  to 


64        Industrial  Organic  Analysis 

cool  very  slowly  while  it  is  constantly  stirred  with  a 
thermometer,  graduated  in  tenths  of  a  degree  ;  when  the 
solidifying  point  is  reached,  the  thermometer  should  not 
be  falling  more  than  half  a  degree  per  minute.  The 
solidifying  point  is  the  temperature  at  which  a  small 
crystal  of  phenol,  introduced  into  the  liquid,  just  begins 
to  grow.  If  the  cooling  is  carried  too  far,  the  tempera- 
ture rises  again  when  crystallisation  sets  in  ;  the  highest 
point  reached  is  the  real  solidifying  point. 

(b)  Water. — This  constituent  may  be  determined  in 
crystallised  carbolic  acid  by  mixing  it  with  5  times  its 
weight  of  dry  levigated  lead  oxide  and  drying  at  70°  to 
80°  until  constant  in  weight.  The  function  of  the 
lead  oxide  is  a  mechanical  one,  the  object  of  the  admix- 
ture being  to  facilitate  the  evaporation  of  the  water. 

As  little  as  i  per  cent,  of  water  may  be  detected  by  the 
milkihess  produced  when  the  phenol  is  shaken  with  an 
equal  volume  of  chloroform. 

Crude  carbolic  acid,  consisting  either  entirely  of 
cresylic  acid,  i.e.,  cresols,  or  of  a  mixture  of  the  latter 
with  phenol,  is  largely  used  for  disinfecting  purposes. 
Very  often,  cresylic  acid  is  mixed  with  soft  soap  and 
water  to  form  an  emulsion,  owing  to  its  sparing  solu- 
bility in  water.  The  analysis  of  such  preparations  will 
be  described  in  the  chapter  on  soap.  Pure  phenol  is 
used  in  medicine,  and  in  the  manufacture  of  dyes  and 
picric  acid. 

THE  EXAMINATION  OF  CREOSOTE  OIL  AND  ITS  PRODUCTS. 

This  portion  of  the  coal  tar  distillate  is  a  viscous, 
greenish-yellow  fluorescent  liquid  which  partially  solidi- 
fies on  cooling,  owing  to  the  separation  of  naphthalene. 
The  solid  matter  amounts  to  some  20  per  cent.  Creosote 
oil  has  an  average  specific  gravity  of  1-04.  Its  principal 
constituents  are  naphthalene,  which  usually  amounts  to 
20  to  30  per  cent,  or  more,  and  higher  phenols,  including 


Coal  Tar  and  its  Distillation  Products   65 

cresols,  xylenols,  naphthols,  etc.,  which  amount  to  10  to 
20  per  cent.  In  addition  to  these  may  be  mentioned 
aniline,  and  other  basic  compounds  such  as  acridine, 
cryptidines,  and  quinoline,  and  hydrocarbons  such  as 
methyl  naphthalene,  diphenyl,  anthracene,  acenaphtene, 
hydronaphthalenes,  etc. 

On  redistillation,  a  little  light  oil  and  carbolic  acid  are 
separated,  and  the  two  main  fractions  obtained,  i.e.,  first 
and  second  naphthalene  oils,  are  allowed  to  cool,  and  the 
naphthalene  crystallising  out  is  removed  by  nitration 
and  further  purified.  The  mother  liquors  contain 
varying  proportions  of  naphthalene  and  other  hydro- 
carbons, about  10  to  30  per  cent,  of  phenols  and  smaller 
amounts  of  basic  substances.  The  liquors  obtained 
from  the  more  volatile  first  naphthalene  oil  are  used  as 
disinfectants,  sometimes  by  themselves,  sometimes  in 
the  form  of  emulsions  with  soft  soap,  or  mixed  with  lime, 
kieselguhr,  borax,  or  other  solid  material.  They  may 
also  be  redistilled  to  yield  a  product  containing  about 
50  per  cent,  of  phenols  which  is  mixed  with  zinc  chloride 
solution  to  from  an  emulsion  for  pickling  timber.  The 
mother  liquors  form  the  second  naphthalene  oil,  are  also 
largely  used  for  preserving  timber,  i.e.,  harbour  piers, 
railway  sleepers,  etc. 

The  preparations  containing  soap,  such  as  creoline  and 
lysol,  are  treated  of  in  the  chapter  on  Soap.  Under  the 
present  heading  the  analysis  of  phenolic  disinfecting 
powders,  creosoting  liquors  and  the  testing  of  naphtha- 
lene for  purity  are  described. 

THE  EXAMINATION  OF  PHENOLIC  POWDERS  FOR  NEUTRAL 
TAR  OILS  AND  PHENOLS. 

Different  methods  must  be  employed,  (i.)  for  prepara- 
tions made  from  powders  which  do  not  combine  chemi- 
cally with  the  phenols,  such  as  borax  or  kieselguhr,  and 
(ii.)  for  those  made  from  lime  or  other  basic  substances 
forming  salts  with  the  phenols.  As  is  well  known,  the 
phenols  are  neutralised  by  the  oxides  or  hydroxides  of 
the  alkali  metals  or  metals  of  the  alkaline  earths,  but  not 
by  alkaline  carbonates,  borates,  etc. 
I.O.A.  5 


66 


Industrial  Organic  Analysis 


(i.)  Preparations  containing  non- 
basic  Powders. — This,  as  well  as 
the  following  method  for  powders 
with  basic  constituents,  is  taken 
from  Allen's  "  Commercial  Organic 
Analysis  "  (revised  edition)  :— 

50  grams  of  the  powder  are  placed 
in  a  thimble  of  filter  paper  and 
extracted  with  ether  in  a  Soxhlet 
extraction    apparatus,    which    is 
shown  in  the  accompanying  illus- 
tration.   The   three  parts  of   the 
apparatus,  i.e.,  the  flask,  the  ex- 
tractor and  the  condenser,  may  be 
connected  by  corks  or,  better  still, 
by  means  of  ground  glass  joints. 
The  solvent  is  placed  in  the  flask 
below,  and  kept  actively  boiling  on 
the  water  bath  ;  as  it  is  condensed 
above,    it    falls    back    into     the 
extractor     and    is    automatically 
syphoned  off  as  soon  as  it  reaches 
a  certain  level.     In  this  way  the 
material  is  continually  being  ex- 
posed to  the  solvent  action  of  fresh 
ether,  while  the  dissolved  material 
collects  in  the  flask.     The  intro- 
duction of  a  tap  in  the  lower  part 
of   the   syphon  bend  is  a  useful 
modification  due  to  Lewkowitsch  ; 
by  its  means  a  small  portion  of  the 
FIG.    7.— Soxhlet  Ex-  solvent  may  be  withdrawn   from 

traction    Apparatus  tjme   to   tjme   an(j    evaporated   to 

with     Davies     Con- 

denser  dryness  ;   when   no  residue  is  leit, 


Coal  Tar  and  its  Distillation  Products   67 

the  extraction  process  is  known  to  be  at  an  end. 
The  ethereal  solution  which  contains  the  phenol  is 
shaken  with  20  c.c.  of  a  20  per  cent,  solution  of  sodium 
hydroxide.  (The  amount  of  soda  solution  may  be 
varied  according  to  the  amount  of  phenols  supposed  to 
be  present,  using  I  c.c.  for  each  I  per  cent.)  The  ethereal 
solution  is  separated  off,  and  the  alkaline  solution 
shaken  out  with  two  successive  portions  of  ether,  in 
order  to  extract  the  whole  of  the  neutral  tar  oils.  The 
united  ether  solutions  are  then  shaken  out  with  a  small 
quantity  of  sodium  hydroxide  solution,  in  order  to 
remove  any  traces  of  phenols  which  may  remain  with  the 
neutral  oils,  evaporated  in  a  tared  flask  on  the  water 
bath,  and  weighed,  after  drying  at  100°  C.  for  i  hour. 

The  united  alkaline  solutions  are  boiled  down  in  a 
flask  to  about  10  c.c.  in  volume,  transferred  to  a 
graduated  cylinder  or  burette,  and  acidified  with  dilute 
sulphuric  acid  containing  i  part  of  the  concentrated 
acid  to  3  parts  of  water.  When  the  mixture  is  quite 
cold  the  volume  of  the  separated  phenols  is  read  off,  the 
weight  being  estimated  on  the  assumption-  that  i  c.c. 
weighs  1-050  grams.  If  the  presence  of  fatty  or  resin 
acids  is  suspected,  the  contents  of  the  burette  are  trans- 
ferred to  a  flask  and  submitted  to  steam  distillation ; 
the  phenols  will  pass  over  the  distillate,  while  the  acids 
will  remain  in  the  flask.  The  latter  may  be  characterised 
by  their  solubility  in  sodium  carbonate  solution. 

(ii.)  Preparations  containing  Basic  Powders. — 50  grams 
of  the  alkaline  powder  are  mixed  in  a  mortar  with  5  c.c. 
of  water,  after  which  strong  sulphuric  acid  is  added, 
drop  by  drop,  at  intervals  and  mixed  well  into  the  powder 
or  paste  by  means  of  a  pestle  ;  the  addition  of  the  acid 
should  extend  over  several  hours,  in  order  to  avoid  a 
sensible  rise  in  temperature  ;  it  is  continued  until  the 

5—2 


68        Industrial  Organic  Analysis 

whole  mass  is  distinctly  acid.  If  the  resulting  product 
is  a  paste,  it  is  mixed  with  sufficient  sand  to  bring  it  into 
a  granular  form  ;  it  is  then  extracted  with  ether  in  the 
Soxhlet  apparatus,  and  the  extract  is  submitted  to  the 
same  treatment  as  described  above  for  non-alkaline 
powders. 

According  to  Allen,  good  powders  should  contain  from 
12  to  18  per  cent,  of  phenols.  It  is  doubtful  whether 
powders  containing  lime  are  as  efficient  for  disinfecting 
purposes  as  those  in  which  the  phenols  are  present  in  the 
free  state. 

THE  TESTING  OF  CREOSOTING  LIQUOR. 

Most  of  the  creosote  oil  distilled  is  used  for  preserving 
timber  which  is  to  be  used  for  railway  sleepers,  telegraph 
poles,  harbour  piers,  etc.  The  basic  constituents  of  the 
oil  are  believed  to  be  of  as  much  value  as  the  phenols  as 
antiseptics,  and,  moreover,  they  are  said  to  form  resin- 
like  combinations  with  the  phenols  and "  unsaturated 
hydrocarbons,  which  fill  up  the  pores  of  the  wood  and 
protect  it  from  destruction  by  micro-organisms.  Exten- 
sive adulteration  with  lignite,  shale  and  rosin  oils,  which 
contain  no  bases,  is  therefore  objectionable.  The 
naphthalene  and  other  neutral  constituents  are  also 
believed  to  be  beneficial,  for,  being  introduced  into  the 
wood  in  a  liquid  state  at  about  50°,  they  solidify  in 
the  pores  and  form  a  protective  covering  throughout. 

In  testing  creosoting  liquor  it  is  often  necessary  to 
adhere  to  a  method  described  in  a  contract  in  which  are 
stated  the  conditions  which  the  article  must  fulfil. 
With  such  a  complex  mixture  as  creosoting  liquor,  the 
analytical  results  often  vary  considerably  with  slight 
variations  in  the  method  of  analysis,  and  must  therefore 
be  judged  with  due  regard  to  the  process  by  which  they 
have  been  obtained. 

The  foU owing  is  a  condensed  extract  of  Tidy's  specifica- 
tion : — 

(i)  "  The  creosote  oil  must  be  completely  liquid  at 
38°  C.,  no  deposit  taking  place  until  it  is  cooled  to  35°  C." 


Coal  Tar  and  its  Distillation  Products   69 

(This  brings  the  percentage  of  naphthalene  and  other 
solid  hydrocarbons  within  certain  limits.) 

(2)  "  It  must  contain  at  least  25  per  cent,  of  con- 
stituents not  distilling  over  at  (or  below)  316°  C. 

(3)  "  Tested  by  the  process  described  below,  it  shall 
show  a  total  of  not  less  than  8  per  cent,  of  tar  acids 
(phenols) . 

(4)  "  It  shall  contain  no  admixture  of  bone  oil,  shale 
oil,  or  any  substance  not  derived  from  coal  tar." 

The  Determination  of  the  Coal  Tar  Acids  (Phenols). — 
ipo  c.c.  of  the  well  mixed  creosote  is  distilled  at  a  tempera- 
ture of  316°  till  there  is  no  further  distillate.  The 
distillate  is  mixed  and  well  shaken  in  a  stoppered  flask 
with  a  solution  of  sodium  hydroxide  of  specific  gravity 
1-200  ;  the  mixture  is  heated  on  the  water  bath  and  then, 
after  the  stopper  has  been  replaced,  shaken  vigorously  for 
i  minute.  The  contents  of  the  flask  are  poured  into  a 
separating  funnel,  and  the  soda  solution  is  drawn  off 
when  the  layers  have  separated.  The  creosote  is  heated 
a  second  and  a  third  time  with  fresh  portions  of  20  c.c. 
of  the  soda  solution,  the  process  being  repeated  just  as 
described.  The  three  soda  solutions  are  mixed,  and 
when  cold,  any  particles  of  oily  matter  are  got  rid  of  by 
the  use  of  a  separating  funnel  ;  the  solution  is  then 
boiled  vigorously  to  expel  the  last  traces  of  neutral  or 
basic  oil.  On  cooling,  dilute  sulphuric  acid  is  added 
(i  part  acid  to  3  of  water)  until  the  whole  is  slightly 
acid,  and  the  mixture  is  transferred  to  a  separating  funnel 
and  allowed  to  stand  to  cool  until  the  tar  acids  have 
separated.  The  latter  are  now  separated  off  and 
dissolved  in  20  c.c.  of  soda  solution  of  the  same  strength 
as  that  used  before,  and  10  c.c.  of  water ;  the  alkaline 
solution  is  well  cooled,  transferred  to  a  100  c.c.  measure 
and  acidified  with  dilute  sulphuric  acid  (i  to  3),  of  which 
about  35  c.c.  will  be  sufficient.  The  whole  is  allowed 


70        Industrial  Organic  Analysis 

to  stand  for  2  hours  to  cool,  after  which  the  volume  of 
the  tar  acids  is  read  off. 

Instead  of  distilling  the  creosote  to  316°,  Sad  tier 
prefers  to  distil  to  the  point  of  pitching,  and  thus  includes 
some  of  the  higher  phenols  in  the  estimation,  which  are 
rejected  along  with  the  still  residue  in  Tidy's  process  ; 
in  view  of  the  fact  that  the  higher  phenols  are  in  all 
probability  just  as  efficacious  in  the  preserving  of  timber 
as  the  lower  members  of  the  series,  this  modification 
would  seem  to  be  desirable.  Further,  in  extracting  the 
phenols,  Sadtler  recommends  the  use  of  10  per  cent, 
sodium  hydroxide  solution  at  first,  followed  by  successive 
portions  of  30  per  cent,  solution  in  the  subsequent 
extractions,  until  all  the  phenols  have  been  dissolved 
out.  In  order  to  get  a  good  separation,  petroleum 
ether  may  be  added ;  the  separation  of  the  last  traces 
of  neutral  oils  is  also  facilitated  by  the  use  of  this 
solvent,  which  may  be  completely  removed  by  subse- 
quently boiling  the  alkaline  solution. 

Estimation  of  the  Basic  Constituents. — The  following 
method  is  due  to  Sadtler;  100  c.c.  of  the  creosote  oil 
are  distilled  from  a  retort  to  the  point  of  coking,  and  the 
distillate  is  agitated  with  two  successive  portions  of 
20  c.c.  of  dilute  sulphuric  acid  (i  to  3).  The  acid 
solution  which  contains  the  bases  in  the  form  of  sulphates 
is  separated  off  and  rendered  alkaline  by  the  addition  of 
sodium  hydroxide  solution.  The  bases  are  liberated, 
partly  in  the  form  of  an  oily  layer,  and  partly  in  aqueous 
solution.  The  oily  layer  is  separated  off,  and  the 
alkaline  liquid  is  distilled  nearly  to  dryness.  The 
distillate  is  mixed  with  the  oily  extract  and  the  whole  is 
acidified  with  hydrochloric  acid,  and  evaporated  to 
dryness  on  the  water  bath.  The  residue,  which  consists 
of  the  hydrochlorides  of  the  tar  bases,  is  dissolved  in  a 


Coal  Tar  and  its  Distillation  Products    71 

small  quantity  of  water,  and  solid  sodium  hydroxide  is 
dissolved  in  the  liquid  till  a  saturated  solution  is  obtained. 
The  bases  may  then  be  separated,  weighed,  and,  if 
desired,  further  examined  by  conversion  into  platini- 
chlorides.  They  may,  however,  be  more  conveniently 
estimated  by  the  following  titration  process  (Bureau  of 
Animal  Industry,  Bulletin  107,  U.S.  Dept.  of  Agricul- 
ture), which  was  originally  devised  for  the  estimation  of 
tar  bases  in  sheep  dips.  (See  Chapter  IV.) 

The  acid  solution  of  the  bases  is  filtered  if  necessary, 
made  up  to  300  c.c.  with  water,  and  divided  into 
two  equal  portions  in  two  similar  titration  flasks  of 
about  300  c.c.  capacity.  To  the  contents  of  one  of 
these  flasks  is  added  a  drop  or  two  of  methyl  orange 
solution,  and  then  semi-normal  sodium  hydroxide  solution 
until  the  red  tint  just  disappears,  as  nearly  as  can  be 
judged  by  comparison  with  the  other  portion  which  has 
been  treated  with  an  equal  amount  of  methyl  orange. 
O'l  to  0-2  c.c.  of  semi-normal  soda  solution  are  then 
added.  This  first  titration  is  not  quantitative,  but  is 
carried  out  to  obtain  a  standard  by  which  the  second 
portion  may  be  titrated  to  neutrality. 

A  neutral  solution  of  the  chlorides  or  sulphates  of 
the  bases  is  thus  obtained;  any  further  addition 
of  alkali  would  produce  an  alkaline  reaction  with  the 
methyl  orange.  Phenol  phthalein  is  now  added,  and 
titration  is  continued  to  the  end  point  of  this  indicator  ; 
the  reason  why  an  alkaline  reaction  is  not  observed 
immediately  is  that  the  alkali  displaces  the  tar  bases 
from  combination  with  the  mineral  acids,  and  the 
liberated  bases  have  no  action  on  the  phenol  phthalein  ; 
only  when  the  whole  of  the  bases  has  been  liberated 
will  the  alkali  added  produce  a  coloration  with  the 
phenol  phthalein.  The  number  of  cubic  centimetres 


72        Industrial  Organic  Analysis 

of  soda  solution  used  between  the  two  indicators, 
multiplied  by  0-079,  gives  the  total  amount  of  the  bases 
in  terms  of  pyridine,  or,  multiplied  by  0-129, m  terms  of 
quinoline. 

NAPHTHALENE. 

This  substance  occurs  in  coal  tar  to  the  extent  of 
5  to  10  per  cent.,  being  one  of  its  chief  constituents.  In 
the  dyeing  industry  it  is  of  great  importance  as  the  source 
of  the  naphthols,  naphthylamine,  phthalic  acid,  etc.  As 
mentioned  above,  naphthalene  is  obtained  from  the 
carbolic  oil  and  creosote  oil  fractions  of  coal  tar ;  it  is 
freed  from  phenols  by  washing  with  caustic  alkali 
solution,  and  sometimes  with  sulphuric  acid,  which 
removes  the  phenols  as  sulphonic  acids.  Further 
purification  is  effected  by  filter-pressing  the  warm 
material,  and  fractional  distillation. 

The  Testing  of  Naphthalene  for  Purity. — Naphthalene 
which  is  to  be  chemically  treated  is  required  to  be  as 
pure  as  possible.  The  purest  naphthalene  melts  at 
79*5°  to  79-8°  and  boils  at  217°  to  218°  at  760  mm. 
pressure.  The  following  tests  for  purity  may  be  applied 
to  commercial  naphthalene  : — 

(1)  Dissolve  in  hot  pure  concentrated  sulphuric  acid  ; 
the  solution  should  turn  only  faintly  purple  or  pink. 
With  less  pure  brands  it  will  turn  red. 

(2)  Pour  pure  fuming  nitric   acid   on  to  the  bottom 
of  a  desiccator,  and  place  the  naphthalene  in  a  watch 
glass  above  it,  covering  up  the  whole  as  usual  ;   if  the 
sample  remains  white  for  half  an  hour  it  is  good,  and  if 
for  2  hours,  it  is  excellent.     Inferior  qualities  soon  turn 
pink.     After  some  hours,  all  samples  go  yellow,  probably 
owing  to  the  formation  of  nitronaphthalene. 

(3)  Phenols  may  be  tested  for  by  boiling  the  sample 
with  dilute  sodium  hydroxide  solution,  cooling,  filtering 
and  adding  to  the  filtrate  a  little  bromine  water  an4 


Coal  Tar  and  its  Distillation  Products    73 

dilute  hydrochloric  acid ;  any  phenols  which  may  be 
present  will  be  precipitated  as  bromine  derivatives. 

(4)  Quinoline  bases  may  be  tested  for  by  dissolving 
the  sample  in  concentrated  sulphuric  acid,  pouring  the 
solution  into  water  and  filtering  ;  on  making  the  filtrate 
alkaline  and  distilling,  the  bases  will  pass  over  with  the 
steam  and  be  recognisable  by  their  characteristic 
smell. 

For  naphthalene  which  is  to  be  used  as  insecticide  or 
for  carburetting  gas,  the  tests  for  purity  need  not  be  so 
stringent. 

ANTHRACENE  OIL  PRODUCTS. 

Anthracene  oil,  which  is  distilled  from  about  270° 
to  the  point  of  pitching,  is  a  greenish  yellow  fluid,  turning 
brown  on  exposure  to  air,  and  boiling  from  280°  to  400°. 
On  cooling  to  the  ordinary  temperature  it  becomes  semi- 
solid,  yielding  a  crystalline  deposit  of  crude  anthracene, 
amounting  to  about  30  per  cent,  of  the  total.  Crude 
anthracene  is  an  exceedingly  complex  mixture ;  in 
addition  to  many  substances  of  unknown  constitution,  it 
contains  the  following  :  naphthalene  and  homologues, 
anthracene,  methyl  anthracene,  etc.,  phenanthrene, 
acenaphtene,  diphenyl,  pyrene,  fluorene  and  other  hydro- 
carbons of  high  molecular  weight,  phenols  of  complex 
constitution,  and  nitrogen  compounds  such  as  cabazole, 
acridine  and  imidophenyl  naphthyl.  The  fluid  portions 
of  the  anthracene  oil  are  of  little  value  ;  they  are  either 
redistilled  or  used  as  lubricants.  The  solid  constituents 
are  worked  up  for  anthracene,  which  is  one  of  the  most 
valuable  constituents  of  coal  tar,  being  the  basis  of  the 
alizarin  dyes.  It  is  present  in  coal  tar  to  the  extent  of 
about  0-3  to  0-9  per  cent.,  and  constitutes  from  2|  to  3| 
per  cent,  of  the  anthracene  oil.  The  other  constituents 
of  anthracene  oil,  with  the  exception  of  carbazole,  are  of 
little  or  no  commercial  importance. 

After  pressing,  first  cold  and  then  hot,  a  crude  anthra- 
cene cake  containing  some  30  to  40  per  cent,  of  anthra- 


74        Industrial  Organic  Analysis 

cene  is  obtained.  A  second  crop,  crystallised  from  the 
anthracene  oil  at  about  15°,  yields  an  anthracene  cake 
containing  10  per  cent,  of  anthracene.  The  subsequent 
processes  by  which  pure  anthracene  is  obtained  consist 
mainly  in  washing  the  finely  divided  product  with 
solvents  such  as  naphtha,  creosote  oil,  acetone  or  liquid 
sulphur  dioxide. 


THE  EXAMINATION  OF  CRUDE  OR  PURIFIED  ANTHRACENE. 

The  value  of  the  product  depends,  firstly,  on  the  amount 
of  real  anthracene  it  contains,  and  secondly,  on  its  free- 
dom from  undesirable  impurities,  the  chief  of  which  are 
the  higher  paraffins  and  methyl  anthracene,  which  are 
often  present  in  appreciable  quantities  in  some  tars, 
notably  those  from  Scotland  and  the  north  of  England. 
These  impurities  are  very  difficult  to  remove,  and  if 
present  in  appreciable  quantities,  render  the  anthracene 
unfit  for  alizarin  making ;  the  paraffins  especially 
impede  the  oxidation  to  anthraquinone,  and  render  the 
purification  of  the  latter  difficult.  Carbazole  is  another 
objectionable  impurity  which  should  be  removed  as 
completely  as  possible.  The  estimation  of  anthracene 
and  the  detection  and  estimation  of  the  impurities  just 
mentioned  are  described  below. 

Crude  unwashed  anthracene  cake  formerly  contained 
about  30  to  40  per  cent,  of  anthracene,  but  in  recent 
years  it  has  been  found  possible  to  obtain  a  product 
containing  already  at  this  stage  from  40  to  50  per  cent., 
or  even  more,  of  anthracene. 

The  Estimation  of  Anthracene  in  Crude  or  Purified 
Anthracene  Cake. — The  method  usually  adopted  is  known 
as  the  "  Hochst  test,"  in  which  the  anthracene  is 
quantitatively  oxidised  to  anthraquinone,  by  means  of 
chromic  acid  in  acetic  acid  solution,  and  estimated  as 
such.  During  the  process  practically  all  the  accompany- 
ing substances  are  either  completely  oxidised  or  converted 


Coal  Tar  and  its  Distillation  Products    75 

into  products  which  are  easily  removed  by  washing  with 
water  or  dilute  alkaline  solution.  In  order  that  reliable 
results  may  be  obtained,  the  following  directions  should 
be  closely  adhered  to  : — 

i  gram  of  the  carefully  sampled  anthracene  cake  is 
treated  with  45  c.c.  of  pure  glacial  acetic  acid  in  a  500  c.c. 
flask,  fitted  with  a  reflux  condenser,  and  the  mixture  is 
kept  boiling  while  a  solution  of  15  grams  of  pure  chromic 
acid  in  10  c.c.  of  glacial  acetic  acid  and  10  c.c.  of  water 
is  added  drop  by  drop ;  the  addition  of  this  oxidising 
mixture  is  to  extend  over  2  hours,  after  which  the 
boiling  is  continued  for  2  hours  more.  The  mixture  is 
then  allowed  to  stand  for  12  hours,  when  400  c.c.  of  cold 
water  are  added.  After  another  3  hours,  the  precipi- 
tated anthraquinone  is  collected  on  a  filter  and  washed, 
first  with  distilled  water,  then  with  200  c.c.  of  a  boiling 
I  per  cent,  solution  of  potassium  hydroxide,  and  finally 
with  hot  distilled  water.  The  quinone  is  then  transferred, 
with  the  aid  of  a  wash  bottle,  to  a  dish,  dried  at  100°, 
and  treated  with  10  c.c.  of  fuming  sulphuric  acid  at  100° 
for  10  minutes  on  the  water  bath.  The  solution  thus 
obtained  is  poured  into  a  flat  dish  and  kept  for  12  hours 
in  a  moist  atmosphere,  say,  on  a  thick  layer  of  moist 
blotting  paper  under  a  bell  jar,  to  absorb  water.  200  c.c. 
of  cold  water  are  then  added  to  the  contents  of  the  dish  ; 
the  precipitated  anthraquinone  is  collected  on  a  filter, 
and  washed,  first  with  distilled  water,  then  with  boiling 
i  per  cent,  potassium  hydroxide  solution,  and  finally 
with  hot  distilled  water.  The  residue  on  the  filter  is 
transferred  to  a  platinum  dish,  dried  and  weighed  ;  after 
volatilising  the  quinone  at  a  gentle  heat,  the  dish  is 
re-weighed  with  the  particles  of  ash  and  coal  which 
remain.  The  difference  between  the  two  weights  is 
equal  to  the  weight  of  the  anthraquinone,  which, 


76        Industrial  Organic  Analysis 

multiplied  by  0-8558,  is  equal  to  the  weight  of  real 
anthracene  in  the  sample. 

The  anthraquinone  got  by  the  above  process  should 
be  crystalline,  and  of  a  pale  yellow  colour.  An  orange 
or  red  colour  indicates  the  presence  of  the  quinones  of 
other  hydrocarbons,  especially  those  of  phenanthrene  or 
chrysene,  the  latter  being  recognisable  by  the  indigo 
coloration  which  it  produces  on  the  addition  of  con- 
centrated sulphuric  acid.  The  quinone  of  imido  phenyl 
naphthyl  prevents  the  crystallisation  of  the  anthra- 
quinone ;  according  to  Allen,  it  may  be  destroyed  by 
longer  heating  with  fuming  sulphuric  acid.  In  the 
presence  of  methyl  anthracene  the  anthraquinone 
obtained  does  not  show  the  usual  characteristic  needles, 
but  it  is  more  or  less  felted. 

Carbazole,1  like  imido  phenyl  naphthyl,2  produces  a 
quinone  which  interferes  with  the  purification  of  the 
anthraquinone  ;  according  to  Behrens,  it  is  detected  by 
extracting  the  anthracene  with  cold  ethyl  acetate, 
evaporating  the  solvent  on  a  watch  glass  and  warming 
with  a  few  drops  of  nitrobenzene  and  phenanthra- 
quinone  ;  characteristic  narrow  plates  of  a  coppery 
lustre  are  got  if  carbazole  is  present. 


Estimation  of  Carbazole. — This  substance,  which 
occurs  in  coal  tar  to  about  the  same  extent  as  anthracene, 
has  lately  become  of  some  importance  as  raw  material  in 
the  cyanide  industry,  as  well  as  in  the  manufacture  of 
dyes.  Owing  to  the  weakly  acidic  nature  of  the  imido 
group  which  it  contains,  it  may  be  separated  from  the 
accompanying  constituents  by  heating  with  potash, 
when  a  potassium  derivative  is  formed. 


C6H4 CeH*  2  C6H4 CioH6 

NH  NH 


Coal  Tar  and  its  Distillation  Products    77 

For  its  estimation,  Kraemer  and  Spilker  recommend 
the  following  process :  the  crude  anthracene,  in  a 
finely  divided  state,  is  extracted  with  warm  dilute 
sulphuric  acid,  which  removes  all  other  nitrogen  com- 
pounds which  are  of  a  basic  nature.  The  nitrogen  is 
then  estimated  in  the  residue  by  Kjeldahl's  process 
(see  p.  22)  and  calculated  to  carbazole. 

The  Determination  of  Paraffins  in  Crude  Anthracene. — 
Kraemer  and  Spilker  recommend  the  following  process, 
which,  in  theory,  is  the  same  as  their  process  for  estimat- 
ing paraffins  in  benzol  (see  p.  51) :  10  grams  of  the 
finely  powdered  anthracene  are  shaken  with  70  c.c.  of 
ether  in  a  100  c.c.  measuring  flask  for  10  minutes,  the 
flask  is  filled  to  the  mark  with  ether,  and  the  contents 
are  allowed  to  settle.  50  c.c.  of  the  clear  solution, 
representing  5  grams  of  the  sample,  are  introduced  into 
a  porcelain  dish,  the  ether  is  allowed  to  evaporate,  and 
the  residue  is  dried  at  100°  for  half  an  hour.  After 
cooling,  it  is  triturated  with  8  c.c.  of  fuming  sulphuric 
acid,  containing  20  per  cent,  of  SO3.  The  dish  is  covered 
with  a  watch  glass  and  heated  to  100°  for  3  hours,  with 
frequent  stirring.  The  contents  are  then  washed  into 
a  beaker  with  500  c.c.  of  hot  water  and,  after  cooling, 
passed  through  a  dry  filter.  The  beaker  is  rinsed  out 
on  to  the  filter,  which  is  washed  with  water  until  barium 
chloride  solution  no  longer  produces  a  precipitate  in  the 
filtrate ;  the  filter  is  then  allowed  to  drain,  thoroughly 
moistened  with  absolute  alcohol,  and  the  paraffin  is 
washed  by  means  of  ether  into  a  weighed  dish  until  a 
few  drops  of  the  running  ether  leave  no  residue  on 
evaporation.  The  last  traces  of  paraffin  are  removed 
from  the  beaker  by  means  of  ether,  which  is  also  passed 
through  the  filter  and  added  to  the  main  portion.  The 
total  ether  solution  is  evaporated  in  the  dish,  the 


78        Industrial  Organic  Analysis 

residue  dried    at    105°  for  2   hours,   and  weighed  as 
paraffin. 

Good  anthracenes  should  contain  little  or  no  paraffin  ; 
inferior  qualities  may,  however,  contain  about  4  to  6 
per  cent,  of  this  constituent. 


REFERENCES. 

"  Coal  Tar  and  Ammonia,"  by  Lunge.  1910.  Gurney  and 
Jackson,  London. 

"  Technical  Methods  of  Chemical  Analysis,"  Vol.  II.,  Part  II., 
by  Lunge,  translated  by  Keane.  1911.  Gurney  and  Jackson, 
London. 

Allen's  "  Commercial  Organic  Analysis,"  revised  edition. 
1910.  J.  and  A.  Churchill,  London. 

"  Chemisch-Technische  Analyse,"  Vol.  II.,  by  Post.  Fr. 
Vieweg  und  Sohn,  Brunswick. 


CHAPTER  III 

THE   FATTY   OILS   AND   FATS 

INTRODUCTORY. 

THE  fatty  oils  and  fats  constitute,  from  the  chemical 
point  of  view,  a  well  defined  group  of  naturally  occurring 
products.  Consisting  essentially  of  mixtures  of  the  mixed 
or  simple  neutral  glycerol  esters  of  the  aliphatic  acids 
and  other  acids  belonging  to  allied  series,  they  may  be 
hydrolysed  by  the  action  of  caustic  alkali  with  the  forma- 
tion of  free  glycerol  and  the  potassium  or  sodium  salts 
of  the  fatty  acids  ;  when  the  aqueous  solutions  of  these 
salts,  which  are  generally  known  as  soaps,  are  acidified 
with  mineral  acid,  the  bulk  of  the  fatty  acids  are 
precipitated. 

The  acids  which  are  most  commonly  met  with  in  the 
saponification  products  of  the  oils  and  fats  are  as 
follows  :  — 

(1)  The  aliphatic  acids  containing  an  even  number  of 
carbon  atoms  from  butyric  acid,  C4H8O2,  to  arachidic 
acid,  C20H40O2. 

(2)  Unsaturated  open  chain  acids  yielding  dibromo 
additive  compounds,  of  which  the  more  well-known  are 
oleic   acid,  C18H34O2  (CH3(CH2)7CH  =  CH(CH2)7COOH)  ; 
rapic  acid,  Ci8H34O2;  and  erucic  acid, 


(C8H17  -  CH  =  CH(CH2)nCOOH). 

(3)  Unsaturated  acids  yielding  tetrabromo  additive 
compounds,  of  which  one  of  the  most  well-known  is 
linolic  acid,  Ci8H32O2. 

(4)  Acids  yielding  hexabromo  additive  compounds,  of 
which  the  most  well-known   is  linolenic  acid,  CieHsoOa. 


8o        Industrial  Organic  Analysis 

(5)  Acids  yielding  octobromo  additive  compounds, 
e.g.,  clupadonic  acid,  C^H^O^. 

In  addition  to  the  above,  higher  or  lower  aliphatic 
acids  than  those  mentioned,  acids  containing  uneven 
numbers  of  carbon  atoms  in  the  molecule,  cyclic  un- 
saturated  acids  and  hydroxy  acids  may  be  met  with  in 
small  quantities  or  in  isolated  cases.  The  chemical 
constitution  of  many  of  the  acids  obtained  from  the  oils 
and  fats  is,  as  yet,  doubtful. 

By  far  the  greater  number  of  the  fatty  oils  and  fats 
are  complex  mixtures  from  which  it  is  generally  extremely 
difficult  to  isolate  chemical  individuals ;  many  of  the 
methods  for  their  examination  and  identification  are 
therefore  methods  for  determining  the  mean  of  certain 
chemical  or  physical  characteristics  of  their  constituents 
or  of  the  mixtures  of  fatty  acids  which  have  been  obtained 
from  them  by  certain  standard  methods.  Under  this 
heading  come  the  so-called  "  quantitative  reactions  " 
such  as  the  determination  of  the  iodine  absorbing  power 
and  the  saponification  value,  which  are  described  below. 
Chemical  methods  of  a  more  definite  nature  may,  how- 
ever, be  employed  in  certain  cases  ;  as  an  example  may 
be  mentioned  the  examination  of  the  small  amounts  of 
unsaponifiable  matter  which  are  present  in  all  fatty  oils 
and  fats.  The  vegetable  oils  and  fats,  or  at  least  all  the 
more  commonly  known  members  of  this  group,  contain 
small  amounts  of  an  easily  recognised  substance  known 
as  phytosterol,  which  is  absent  from  all  animal  fats ; 
the  latter,  on  the  other  hand,  contain  small  amounts  of 
cholesterol,  which  is  absent  from  the  vegetable  pro- 
ducts. Another  example  is  seen  in  the  isolation  of 
arachidic  acid  from  arachis  oil.  The  colour  tests,  such 
as  those  employed  for  the  recognition  of  sesame  and 
cotton-seed  oils,  depend  on  the  detection  of  small 
amounts  of  characteristic  constituents  of  these  oils  by 
chemical  means. 

All  fatty  oils  and  fats  naturally  contain  smaller  or 
larger  amounts  of  free  fatty  acids,  presumably  produced 
by  the  partial  hydrolysis  of  the  glycerides .;  in  oils  and 
fats  of  animal  origin  the  proportion  of  free  acids  is 
normally  very  small ;  in  vegetable  oils  and  fats  it  is 


The  Fatty  Oils  and  Fats          81 

generally  larger,  an  extreme  case  being  that  of  palm  oil, 
which  may  contain  from  10  to  80  per  cent,  of  free  fatty 
acids.  One  of  the  principal  objects  of  the  refining  pro- 
cesses to  which  vegetable  oils  and  fats  are  submitted  is 
the  removal  of  these  free  acids  by  treatment  with  limited 
amounts  of  alkali ;  such  processes  are,  however,  only 
undertaken  in  the  case  of  products  in  which  the  amounts 
of  free  acids  are  normally  confined  within  reasonable 
limits. 

All  fatty  oils  and  fats,  especially  those  of  vegetable 
origin,  develop  acidity  on  keeping,  unless  protected  from 
light,  air  and  moisture  ;  in  the  opinion  of  Lewkowitsch, 
this  is  in  a  large  measure  to  be  ascribed  to  the  action  of 
hydrolysing  enzymes  which  become  active  under  certain 
conditions.  The  vegetable  oils  and  fats  are  usually 
stored  as  reserve  food  material  in  seeds  and  fruits,  and 
are  accompanied  by  enzymes,  the  function  of  which 
is  to  convert  them  into  soluble  material  which  can 
be  assimilated  by  the  growing  plant.  Lewkowitsch 
ascribes  the  development  of  rancidity  to  the  com- 
bined action  of  air  and  light  on  the  free  acids  so 
produced. 

Classification  of  the  Fatty  Oils  and  Fats. — Complex 
naturally  occurring  substances  such  as  the  fatty  oils  and 
fats  cannot,  of  course,  be  submitted  to  any  hard  and 
fast  rigid  system  of  classification.  The  system  adopted 
by  Lewkowitsch,  given  below,  is  based  partly  on  the 
origin  of  the  product  and  partly  on  the  drying  power, 
i.e.,  the  power  of  absorbing  oxygen  with  the  formation 
of  more  or  less  viscous  products.  This  important 
property  depends  on  the  proportion  of  unsaturated  acid 
radicles  present  as  well  as  on  the  degree  of  unsaturation 
of  the  latter ;  the  combined  effect  of  these  chemical 
characteristics  is  accurately  measured  by  iodine  absorbing 
power,  or,  as  it  is  generally  called,  the  "  iodine  value  " 
of  the  oil  or  fat.  As  pointed  out  by  Lewkowitsch, 
the  iodine  value  is  the  most  convenient  constant 
on  which  to  base  a  system  of  classification,  as  on  it 
depend  to  a  large  extent  important  physical  proper- 
ties, notably  the  consistence,  melting  and  solidifying 
points. 

I.O.A.  6 


82        Industrial  Organic  Analysis 

The  following  scheme  of  classification  should  be  com- 
pared with  the  table  of  constants  and  characteristics 
given  on  p.  1 06  : — 

i. — Liquid  Fats. 

(A)  Vegetable  oils. 

(a)  Drying,  e.g.,  linseed  and  hempseed  oils. 

(b)  Semi-drying,   e.g.,  soya  bean,   cotton  seed, 

rape  seed,  and  sesame  oils. 

(c)  Non-drying,  e.g.,  arachis,  olive  and  castor  oils, 

(B)  Animal  oils. 

(1)  Marine  animal  oils. 

(a)  Fish  oils,  e.g.,  menhaden  oil. 

(b)  Liver  oils,  e.g.,  cod  liver  oil. 

(c)  Blubber  oils,  e.g.,  seal  and  whale  oils. 

(2)  Terrestrial  animal  oils,  e.g.,  sheep's  foot  and 

neat's  foot  oils. 

2. — Solid  Fats. 

(A)  Vegetable  fats,  e.g.,  palm  oil,  cacao  butter,  palm 
nut  and  cocoanut  oils. 

(B)  Animal  fats. 

(a)  Drying  fats. 

(b)  Semi-drying  fats. 

(c)  Non-drying  fats,  e.g.,  beef  tallow,  mutton 

tallow,  butter  fat. 

The  various  groups  in  this  system  of  classification  are, 
of  course,  by  no  means  sharply  defined,  for  all  gradations 
in  drying  power  and  consistence  are  met  with  amongst 
the  oils  and  fats. 

THE  ESTIMATION  OF  FATTY  MATTER  IN  SEEDS,  OIL- 
CAKE, ETC. 

Vegetable  oils  and  fats  are  obtained  ^rom  seeds,  fruits, 
etc.,  by  submitting  the  disintegrated  material  to  high 
pressure,  either  hot  or  cold,  the  finer,  edible  fats  usually 
being  expressed  at  the  ordinary  temperature.  In  recent 
years  processes  involving  the  extraction  of  fatty  material 
by  means  of  organic  solvents  have  been  introduced. 


The  Fatty  Oils  and  Fats          83 

The  exhausted  material,  known  as  oilcake,  which  usually 
contains  from  6  to  16  per  cent,  of  fat  and  varying  pro- 
portions of  nitrogenous  matter,  is  largely  used  as  fodder 
for  cattle. 

The  method  to  be  described  is  generally  applicable  for 
the  determination  of  fatty  matter  in  oleaginous  seeds 
or  fruits  such  as  rape  seed,  linseed,  soya  beans,  copra 
(i.e.,  dried  cocoanut  endosperm),  etc.,  oilcake,  flour  and 
similar  food  materials. 

If  the  material  contains  much  water,  it  must  first  be 
dried  at  a  gentle  heat ;  complete  drying  is  only  necessary 
if  ether  is  to  be  used  as  the  extracting  solvent  (see 
below) ;  if  any  of  the  other  solvents  mentioned  below  are 
to  be  used  a  rough  drying  will  be  sufficient.  For  the 
determination,  100  grams  of  material  may  be  taken.  It 
is  prepared  for  extraction  with  an  organic  solvent  in  the 
Soxhlet  apparatus  (see  p.  66),  by  disintegrating  in  a 
food-mincing  machine  or  hand  mill,  and  then  triturating 
with  ignited  sand  in  a  mortar,  in  order  that  the  vegetable 
tissue  may  be  thoroughly  broken  up.  If,  however,  the 
material  contains  so  much  fat  that  it  forms  a  pasty  mass 
during  the  latter  process,  it  should  be  submitted  to  a 
preliminary  extraction  in  the  Soxhlet  apparatus  after 
having  been  roughly  broken  up  in  the  mincing  machine 
or  mill,  then  triturated  with  sand  in  a  mortar,  and 
re-extracted.  The  following  solvents  may  be  used  for 
extracting  the  fat :  ether,  petroleum  ether  which  is 
completely  volatile  below  80°,  carbon  tetrachloride  or 
trichloro  ethylene.  It  should  be  noted  that  as  ether  is 
miscible  to  a  certain  extent  with  water,  it  may  extract 
constituents  other  than  fat  unless  the  material  has  been 
thoroughly  dried.  The  other  solvents  mentioned  have 
not  this  disadvantage.  The  two  last  mentioned  solvents 
have  the  advantage  of  being  non-inflammable  ;  trichloro 

6—2 


84        Industrial  Organic  Analysis 

ethylene,  however,  is  especially  liable  to  decompose  with 
evolution  of  hydrochloric  acid  if  exposed  to  light  for 
any  length  of  time  in  presence  of  traces  of  moisture. 

On  the  whole,  therefore,  petroleum  ether  may  be 
regarded  as  the  most  suitable  extracting  medium  for  the 
present  purpose. 

The  process  of  extraction  will  require  several  hours ; 
if  great  accuracy  is  desired,  it  may  often  be  advisable  to 
triturate  the  exhausted  material  with  sand  a  second  time 
and  submit  it  to  further  treatment  in  the  Soxhlet 
apparatus.  The  fat  in  the  flask  is  freed  from  the  solvent 
by  distillation  and  heating  to  105°  in  an  oven,  and 
weighed. 

The  crude  fatty  matter  thus  obtained  may  be  analysed 
for  its  content  of  free  fatty  acids  as  described  under 
""  Acid-  Value,"  and  compared  in  this  respect  with  a 
refined  sample  of  the  same  product.  If  it  is  required  for 
further  examination,  such  as  the  determination  of  its 
iodine  value,  especial  care  must  be  taken  to  free  it  com- 
pletely from  the  solvent  used  in  the  extraction  process, 
at  the  same  time  avoiding  prolonged  exposure  to  air  at  an 
elevated  temperature  ;  the  latter  treatment  may  give 
rise  to  appreciable  oxidation  of  fats  having  high  iodine 
values.  After  most  of  the  solvent  has  been  removed  by 
distillation,  the  fat  may  be  heated  under  reduced  pressure 
to  the  temperature  of  boiling  wafer  till  all  traces  of  the 
solvent  and  water  have  been  removed. 

THE  EXAMINATION  OF  THE  FATTY  OILS  AND  FATS. 

Preparation  of  the  Sample. — Before  the  sample  is  sub- 
mitted to  a  chemical  and  physical  investigation,  it  must 
be  freed  from  foreign  matter  such  as  water,  dissolved 
soaps,  vegetable  or  animal  tissues,  etc.  In  most  cases 
it  will  be  sufficient  to  melt  the  sample,  and,  if  water  is 


The  Fatty  Oils  and  Fats          85 

seen  to  be  present,  allow  it  to  stand  in  a  warm  place  till 
separation  has  taken  place,  when  the  clear  fat  is  passed 
through  a  dry  filter.  Dissolved  mineral  matter  such  as 
soap  is  detected  by  burning  off  a  portion  of  the  sample 
and  examining  the  residue,  if  any,  for  metals  ;  it  may  be 
removed  by  extracting  the  melted  fat  with  dilute  nitric 
acid  and  after  washing  with  warm  water,  proceeding  as 
directed  above.  Foreign  matter  such  as  paraffin,  rosin 
oil,  etc.,  is  not  removed  at  this  stage,  and  is  only 
detected  on  closer  examination. 

The  sample  should  always  be  thoroughly  liquefied,  at  a 
gentle  heat  if  necessary,  and  well  mixed  when  portions 
of  it  are  to  be  abstracted  for  analysis  or  physical  deter- 
minations ;  it  will  be  found  that  oils  often  deposit  small 
quantities  of  solid  matter  at  the  ordinary  temperature, 
which  sink  to  the  bottom  of  the  vessel,  while  solid  fats, 
when  cooled  slowly,  first  deposit  their  higher  melting 
glycerides,  so  that  the  composition  of  the  solidified  mass 
may  not  be  entirely  uniform. 

Estimation  of  Water  in  Fats. — This  determination  is 
of  especial  importance  in  the  case  of  butter  and  margarine, 
in  which  the  water  and  fats  are  mixed  so  as  to  form 
emulsions.  The  determination  may  be  carried  out  as 
follows  :  about  20  grams  of  dried  pumice  in  small  lumps 
are  weighed  in  a  flat  dish,  and  about  10  grams  of  a  fair 
average  sample  are  added.  The  whole  is  then  weighed 
and  dried  in  an  air  oven  at  110°  till  constant  in  weight. 
The  drying  will  take  about  2  hours,  and  should  not  be 
unduly  prolonged,  as  an  increase  in  weight  will  then  take 
place  owing  to  the  oxidation  of  the  fat.  The  loss  in 
weight  of  the  sample  represents  the  water. 

According  to  the  Sale  of  Butter  Regulations  of  1902, 
the  percentage  of  water  in  butter  is  limited  to  16  per  cent. 
Similarly,  margarine  containing  more  than  16  per  cent. 


86        Industrial  Organic  Analysis 

of  water  would  be  held  to  be  not  genuine  margarine  but 
margarine  and  water. 

THE    PHYSICAL    AND    CHEMICAL    CONSTANTS    OF    THE 
FATTY  OILS  AND  FATS. 

Before  describing  the  methods  for  determining  the 
physical  and  chemical  constants  by  which  the  various 
oils  and  fats  may  be  recognised  and  quantitatively 
estimated,  it  should  be  pointed  out  that  these  constants, 
which  are  set  out  for  the  more  well-known  members  of 
the  group  on  p.  106,  are  not  in  the  strict  sense  of  the 
term  constant  for  different  samples  of  the  same  kind  of 
oil  or  fat,  but  are  liable  to  vary  to  a  certain  extent  with 
such  factors  as  the  age,  the  race  or  strain  of  the  animal 
or  plant  from  which  they  were  obtained,  general  treat- 
ment, nature  of  food  or  soil,  climatic  conditions,  etc. 
Having  regard  to  these  variations,  which  will  generally 
not  be  large  enough  to  cause  any  serious  error  in  the 
cases  to  be  dealt  with  here,  it  will  be  realised  that  a 
certain  amount  of  caution  must  always  be  exercised 
in  basing  conclusions  on  the  results  of  the  determination 
of  such  constants  as  the  iodine  and  saponification  values, 
specific  gravity,  etc.  The  acid  value  will  be  influenced, 
not  only  by  the  factors  just  mentioned,  but  also  by  the 
age  and  previous  history  of  the  particular  sample  under 
examination.  The  amount  of  unsaponifiable  matter 
will,  of  course,  be  largely  influenced  by  the  addition  of 
such  substances  as  paraffin. 

Some  of  the  more  important  methods  for  examining 
fatty  oils  and  fats  are  described  below,  after  which 
follow  a  number  of  examples  to  illustrate  how  these 
methods  may  be  applied  in  identifying  the  commoner 
fatty  oils  and  fats,  and  analysing  simple  mixtures  of  the 
latter. 

PHYSICAL  METHODS  FOR  EXAMINING  FATTY  OILS  AND 
FATS. 

The  physical  methods  available  for  the  examination 
of  fatty  oils  and  fats  will  only  be  briefly  dealt  with,  as 


The  Fatty  Oils  and  Fats         87 

more  definite  information  can  generally  be  obtained  by 
the  use  of  chemical  methods.  Useful  indications  may 
sometimes  be  obtained  from  the  specific  gravity,  melting 
point,  refractive  index,  and  sometimes  also  the  optical 
rotation.  Methods  for  determining  the  two  first- 
mentioned  constants  are  described  below,  and  also  the 
"  titer  test "  or  solidifying  point  of  the  mixed  fatty  acids 
derived  from  the  fat.  The  latter  test  is  generally  of 
greater  value  for  purposes  of  characterisation  than  the 
determination  of  the  melting  point  of  the  fat  itself. 

(a)  Specific    Gravity. — This    determination    may    be 
carried  out  by  means  of  the  Sprengel  pyknometer  or  trie 
ordinary  specific  gravity  bottle  ;  in  most  cases,  however, 
sufficiently  accurate  results  may  be  obtained  by  use  of 
the  Westphal  balance  which  is  described  on  p.  42.     In 
case  of  oils  which  are  completely  liquid  at  the  ordinary 
temperature,  the  specific  gravity  is  taken  at  15-5°,  and 
in  case  of  solid  fats,  at  the  temperature  of  boiling  water. 
If  it  is  only  necessary  to  heat  the  fat  a  few  degrees  above 
15-5°  for  complete  liquefaction,  the  determination  may 
be  made  at  this  temperature,  and  the  result  reduced  to 
15-5°.     Allen  gives  the  mean  temperature  correction  for 
most  common  fatty  oils  and  fats,  with  the  exception  of 
whale    oil,    as    0-00064   per  degree   Centigrade.      The 
specific  gravities  of  the  common  oils  and  fats  are  set 
out  in  the  table  on  p.  106.     In  some  cases  the  specific 
gravity   may   furnish   useful    indications   in  detecting 
adulterations. 

For  more  complete  descriptions  of  methods  for 
determining  the  specific  gravities  of  oils  and  fats  see 
the  works  mentioned  at  the  end  of  this  chapter. 

(b)  Melting  Point. — The  methods  commonly  used  for 
relatively  pure  organic  substances  cannot  be  applied  to 
fats.     Numerous  methods  have  been  designed  for  deter- 
mining the  melting  points  of  fats,  and  as  varying  results 
are  obtained  according  to  the  method  used,  it  is  not 


88        Industrial  Organic  Analysis 

advisable  to  place  too  much  reliance  on  indications 
afforded  by  comparison  of  results  obtained  with  figures 
given  in  the  literature.  The  following  method,  elabo- 
rated by  Blichfeldt  for  the  testing  of  edible  fats  and 
fatty  mixtures,  may  be  recommended  as  being  expe- 
ditious and  capable  of  yielding  trustworthy  results  if 
properly  carried  out. 

Into  one  end  of  a  glass  tube,  7  cm.  long,  i  mm.  bore 
and  2  mm.  outside  diameter,  is  introduced  a  column  of 
the  melted  fat,  I  cm.  long.  The  fat  is  suddenly  solidified 
by  placing  the  tube  between  two  flat  pieces  of  ice,  so  that 
the  length  containing  the  fat  is  well  covered.  After 
having  been  kept  on  ice  for  at  least  two  hours,  so  that 
the  fat  may  have  thoroughly  set,  the  tube  is  placed  in  a 
water  bath  so  that  the  top  of  the  fat  column  comes  I  cm. 
below  the  surface ;  while  the  water  is  kept  well  stirred, 
the  temperature,  as  indicated  by  a  delicate  thermometer, 
is  raised  at  the  rate  of  about  i°  per  minute  ;  the  tem- 
perature at  which  the  fat  becomes  sufficiently  soft  to  be 
forced  up  the  tube  by  hydrostatic  pressure  is  taken  as 
the  melting  point.  If  only  one  or  two  melting  points  are 
to  be  determined,  the  tubes  may  be  attached  to  the 
thermometer  bulb  by  a  rubber  band.  If  a  number  of 
tests  are  to  be  carried  out  simultaneously,  the  tubes 
may  be  held  in  position  in  shallow  grooves  in  a  long  strip 
of  wood  by  means  of  a  wide  rubber  band  which  is  wired 
down  at  several  points  throughout  the  length  of  the  strip. 
The  temperature  of  the  water  bath  must  be  raised 
gradually  and  evenly,  the  heat  being  supplied  not  by  a 
direct  flame,  but  by  immersing  in  a  second  water  bath 
kept  at  a  slightly  higher  temperature.  The  stirring 
should  be  carried  out  so  as  to  cause  a  constant  current 
of  water  to  circulate  past  the  melting  point  tubes  and 
thermometer. 

Many  fats  show  perfectly  sharp  melting  points  by  the 


The  Fatty  Oils  and  Fats          89 

above  methods,  duplicate  tests  agreeing  within  0-2°,  or 
at  the  most  0-3°.  In  the  case  of  some  animal  fats  or 
mixtures  of  these  with  oil,  especially  where  the  proportion 
of  stearin  to  oil  is  small,  the  melting  points  observed  are 
less  sharp,  the  movement  of  the  fat  up  the  tube  being 
sluggish. 

(c)  The  Titer  Test. — By  the  titer  test  is  understood 
the  determination  of  the  solidifying  point  of  the  insoluble 
fatty  acids,  together  with  the  small  amount  of  naturally 
occurring  unsaponifiable  matter,  as  obtained  by  the 
process  now  to  be  described.  The  procedure  set  out 
is  that  recommended  by  Lewkowitsch. 

Saponify  100  grams  of  the  oil  or  fat  by  boiling  with 
40  c.c.  of  aqueous  potassium  hydroxide  solution  of 
specific  gravity  1-4,  and  410  c.c.  of  alcohol,  in  a  porcelain 
dish  on  a  water  bath,  stirring  continually  until  the  soap 
becomes  pasty.  Dissolve  the  residual  soap  in  1,000  c.c. 
of  water  and  boil  until  all  the  alcohol  has  been  removed, 
replacing  the  water  which  has  evaporated,  from  time 
to  time.  Decompose  the  soap  solution  by  acidification 
with  dilute  sulphuric  acid  and  when,  by  continued 
boiling,  the  fatty  acids  have  been  obtained  as  a  clear 
layer  floating  on  the  aqueous  liquid,  draw  off  the  latter 
by  means  of  a  syphon,  and  wash  the  fatty  acids  several 
times  with  hot  distilled  water  until  all  acid,  as  tested  for 
by  methyl  orange,  has  been  removed.  Place  the  dish 
with  the  fatty  acids  on  the  water  bath  until  the  latter 
are  melted,  and  the  water  and  impurities  have  settled 
out ;  after  passing  through  a  fluted  filter  in  a  heated 
funnel,  they  will  be  sufficiently  dry  for  further  examina- 
tion. The  actual  determination  may  be  carried  out  by 
the  following  method,  due  to  Dalican  and  recommended 
by  Lewkowitsch : — 

The  mixture  of  fatty  acids,  after  standing  overnight  in 


90        Industrial   Organic  Analysis 

a  desiccator,  is  carefully  melted  in  an  air  bath,  and  as 
much  of  it  is  poured  into  a  test  tube  16  cm.  long  and 
3-5  cm.  wide  as  will  fill  the  tube  somewhat  more  than 
half  full.  The  tube  is  fitted  by  means  of  a  cork  into  a 
wide  mouth  bottle,  10  cm.  wide  and  13  cm.  high  ;  and  an 
accurately  standarised  thermometer,  graduated  in  tenths 
of  a  degree,  from  o°  to  60°  C.,  having  a  mercury  bulb, 
3  cm.  long  and  6  mm.  in  diameter,  is  placed  so  that  the 
bulb  is  in  the  centre  of  the  mass  of  the  fatty  acids. 
When  a  few  crystals  appear  at  the  bottom  of  the  tube, 
the  mass  is  stirred  by  giving  the  thermometer  a  rotatory 
movement,  first  three  times  from  right  to  left  and  then 
three  times  from  the  left  to  right.  The  mass  is  then 
stirred  continually  with  a  quick  circular  movement  of  the 
thermometer  without  allowing  it  to  touch  the  sides  of 
the  vessel,  and  taking  care  that  all  the  solidified  por- 
tions, as  long  as  they  form,  are  well  stirred  into  the  mass 
until  it  has  become  cloudy  throughout.  At  this  point 
the  thermometer  is  carefully  observed ;  at  first  the 
mercury  will  fall  or  remain  stationary,  after  which  it 
will  suddenly  rise-  some  tenths  of  a  degree,  remain 
stationary  for  a  short  time  and  then  fall  again.  The 
maximum  temperature  attained  during  this  rise  is  the 
titer  or  solidifying  point  of  the  mixed  fatty  acids. 

The  above  method  is  stated  by  Lewkowitsch  to  give 
very  concordant  results  ;  it  follows,  of  course,  that 
attention  must  be  paid  to  detail,  especially  in  the  actual 
determination  of  the  solidifying  point. 

The  titer  test  is  largely  used  for  the  commercial  valua- 
tion of  fats  used  in  soap  and  candle  making,  notably 
tallow  and  palm  oil.  It  is  customary  to  stipulate  that 
the  solidifying  point  shall  not  lie  below  a  certain  value, 
say,  43-5°  C.  in  the  case  of  beef  tallow,  and  to  reject  the 
material  or  make  a  deduction  in  the  price  paid  for  it  if  it 
fails  to  comply  with  this  standard. 


The  Fatty  Oils  and  Fats          91 

Reference  to  the  table  on  p.  106  will  show  that  the 
mixed  fatty  acids  derived  from  various  oils  and  fats  may, 
in  some  cases,  readily  be  distinguished  from  one  another 
by  means  of  the  titer  test ;  thus,  compare  the  figures 
given  for  cotton  seed  oil  and  rape  seed  oil.  This 
circumstance  may  be  made  use  of  in  the  examination  of 
fatty  oils  and  fats  in  mixtures  with  mineral  oils,  it  being 
possible  to  separate  the  fatty  acids  from  such  mixtures, 
but  not  the  fatty  oils  or  fats  themselves.  This  problem 
will  be  further  discussed  under  the  examination  of 
lubricating  oils.  (Chapter  V.) 


CHEMICAL  METHODS  FOR  EXAMINING  FATTY  OILS  AND 

FATS. 

Under  this  heading  will  be  described  the  determination 
of  (a)  the  iodine  value,  (b)  the  saponification  value,  and 
(c)  the  Reichert-Wollny  value.  These  may  be  considered 
as  being,  within  certain  limits,  constants  for  the  indi- 
vidual fatty  oils  and  fats,  and  may  often  be  employed  as 
a  basis  for  their  identification  and  approximate  estima- 
tion in  mixtures.  In  addition  will  be  described  the 
determination  of  (d)  the  acid  value,  and  (e)  the  unsaponi- 
fiable  matter.  The  former  of  these  may,  as  explained 
above,  vary  with  the  age  and  previous  history  of  the 
sample,  and  cannot  be  looked  on  as  a  constant  to  be  used 
for  the  purposes  of  identification.  The  amount  of 
unsaponifiable  matter  is  normally  small  in  most  of  the 
common  fatty  oils  and  fats,  and  its  quantitative  deter- 
mination is  generally  of  no  great  importance,  except  in 
cases  where  the  presence  of  added  foreign  material  is 
suspected.  The  qualitative  examination  of  this  con- 
stituent, also  to  be  described  below,  is,  however,  of  some 
importance  in  the  differentiation  of  animal  and  vegetable 
fats,  and  especially  in  the  detection  of  the  latter  as 
adulterants  in  the  former.  (See  "  Phytosteryl  Acetate 
Test.") 

(a)  Iodine  Value. — The  iodine  value  expresses  the 
number  of  parts  by  weight  of  iodine  which  can  be 
absorbed  by  100  parts  of  the  fat.  In  Hiibl's  original 


92        Industrial  Organic  Analysis 

process  the  fat  was  treated  with  iodine  in  an  alcoholic 
solution  containing  mercuric  chloride.  In  Wijs'  process, 
which  is  described  below,  a  solution  of  iodine  mono- 
chloride  in  glacial  acetic  acid  is  used,  the  time  required 
for  absorption  being  considerably  shorter  than  in  Hiibl's 
process. 

In  the  determination  a  weighed  quantity  of  the  fat, 
dissolved  in  carbon  tetrachloride,  is  treated  at  the  ordi- 
nary temperature  with  a  definite  volume  of  the  iodine 
monochloride  solution,  and  an  equal  volume  of  the  latter 
is  set  aside  at  the  same  time,  under  similar  conditions, 
as  a  blank  test.  After  the  requisite  time  has  elapsed 
the  iodine  monochloride  is  estimated  in  each  case  by 
titration  with  sodium  thiosulphate  solution.  The  differ- 
ence in  the  number  of  c.c.  required  in  the  blank  test  and 
in  the  actual  determination  is  calculated  to  express  the 
percentage  proportion  of  iodine  absorbed  by  the  fat. 

The  following  solutions  will  be  required  : — 

Iodine  Monochloride  Solution. — A  10  gram  tube  of 
iodine  trichloride  and  in  grams  of  iodine  are  separately 
dissolved  in  pure  glacial  acetic  acid,  avoiding  absorption 
of  moisture.  The  two  solutions  are  mixed  and  made  up 
to  1,500  c.c.  with  pure  glacial  acetic  acid  (Kahlbaum's 
99  to  100  per  cent,  may  be  used).  The  resulting  iodine 
monochloride  solution  must  be  preserved  in  well- 
stoppered  bottles  to  avoid  access  of  moisture. 

Sodium  Thiosulphate  Solution. — 24  grams  of  the 
crystallised  salt  are  dissolved  in  water  and  made  up  to 
1,000  c.c.  This  solution  may  be  conveniently  standard- 
ised by  the  following  method,  due  to  Volhardt :  3-8631 
grams  of  pure  potassium  dichromate,  free  from  the 
sodium  salt,  are  dissolved  in  water  and  made  up  to 
1,000  c.c.  10  c.c.  of  a  10  per  cent,  solution  of  potassium 
iodide  solution  and  5  c.c.  of  hydrochloric  acid  are  placed 


The  Fatty  Oils  and  Fats          93 

in  a  stoppered  bottle  and  exactly  20  c.c.  of  the  dichro- 
mate  solution  are  run  in.  The  resulting  mixture  now 
contains  exactly  0-2  gram  of  free  iodine,  which  may  be 
titrated  with  the  sodium  thiosulphate  solution,  using 
starch  as  an  indicator,  towards  the  end  of  the  process. 
The  iodine  equivalent  of  the  thiosulphate  solution  may 
then  be  calculated. 

Starch  Solution. — About  I  part  of  starch  is  stirred  up 
in  100  parts  of  water,  and  the  mixture  heated  to  boiling. 

Potassium  Iodide  Solution. — A  10  per  cent,  solution  of 
the  pure  salt  in  water. 

To  determine  the  iodine  value,  the  fat  is  weighed  out 
in  a  small  specimen  tube  which  is  then  dropped  into  a 
narrow  mouthed  bottle  of  about  500  c.c.  capacity,  and 
furnished  with  a  well-fitting  stopper.     The  amount  of 
fat  to  be  taken  varies  with  its  power  of  absorbing  iodine  ; 
in  the  case  of  a  drying  oil,  about  0-15  gram  should  be 
taken,  but  with  a  solid  fat,  having  a  low  iodine  value, 
such  as  cocoanut  oil,  the  amount  may  be  increased  to 
about  1-5  gram.     In  any  case,  the  amount  of  fat  taken 
should  be  so  regulated  that  the  excess  of  iodine  mono- 
chloride   present  after  absorption  is  complete  is  more 
than  half  of  the  amount  originally  present.     The  fat  is 
dissolved  in  10  c.c.  of  carbon  tetrachloride,  and  25  c.c.  of 
the  iodine  monochloride  solution  are  run  in  from  a 
pipette  ;  in  measuring  out  the  solution  for  the  blank  test 
and  for  any  other  determinations  which  are  to  be  made 
simultaneously,  the  same  pipette  should  be  used  and 
allowed  to  empty  itself  in  the  same  manner  and  drain  for 
the  same  length  of  time  in  each  case.     The  resulting 
mixture  should  be  a  clear  solution  ;    if  turbid,  more 
carbon  tetrachloride  should  be  added  till  the  fat  is  com- 
pletely dissolved.     The  bottle  is  now  stoppered  and 
allowed  to  stand  in  a  dark  place  for  the  requisite  time  ; 


94        Industrial  Organic  Analysis 

loss  of  iodine  by  volatilisation  may  be  guarded  against 
by  moistening  the  stopper  with  potassium  iodide  solu- 
tion. The  time  required  for  complete  absorption  varies 
from  ^  an  hour  in  the  case  of  solid  fats  to  2  hours  in  the 
case  of  drying  oils. 

The  estimation  of  the  excess  of  iodine  chloride  is  carried 
out  as  follows  :  20  c.c.  of  potassium  iodide  solution  and 
about  300  c.c.  of  water  are  added  ;  the  mixture  is  then 
titrated  with  the  standard  sodium  thiosulphate  solution, 
shaking  occasionally,  in  order  that  all  the  iodine  may  be 
extracted  from  the  lower  layer  of  carbon  tetrachloride  ; 
a  little  starch  solution  is  added  towards  the  end  of  the 
titration.  The  difference  in  the  number  of  c.c.  of 
thiosulphate  solution  required  in  the  blank  test  and  in 
the  actual  determination  is  then  calculated  to  express 
the  percentage  proportion  of  iodine  absorbed  by  the  fat. 
(The  molecule  of  iodine  chloride,  IC1,  is  chemically 
equivalent  to  the  molecule  of  iodine  I2,  in  the  inter- 
actions involved  in  the  process.) 

The  determination  of  the  iodine  value  is  of  great 
importance  as  a  method  of  characterising  the  fatty  oils 
and  fats.  Owing  to  the  comparatively  low  melting 
points  of  the  glycerides  of  the  unsaturated  acids,  there  is 
a  general  tendency  for  oils  and  fats  with  high  iodine 
values  to  possess  lower  melting  points  and  a  softer  con- 
sistency than  those  with  lower  iodine  values.  This  rule 
is,  of  course,  sometimes  modified  by  the  nature  of  the 
saturated  acid  radicles,  a  case  in  point  being  that  of 
cocoanut  oil,  which  contains  an  unusually  large  pro- 
portion of  saturated  acid  radicles  of  comparatively  low 
molecular  weight,  and  therefore  possesses  a  rather  low 
melting  point  in  spite  of  its  exceptionally  low  iodine 
value. 

As  was  pointed  out  above,  the  drying  power  of  oils 
and  fats  is,  generally  speaking,  directly  proportionate 
to  the  iodine  value.  This  property  of  absorbing  oxygen 


The  Fatty  Oils  and  Fats          95 

on  exposure  to  air  at  the  ordinary  temperature  or  on 
"  blowing  "  with  air  at  elevated  temperatures,  with  the 
formation  of  more  or  less  viscous  products,  finds  exten- 
sive technical  application.  Thus  linseed  oil  (note  the 
high  iodine  value),  especially  after  it  has  been  treated 
with  lead  and  manganese  oxides  at  about  150°, 
readily  dries  to  a  tough  skin  when  exposed  to  air  in  thin 
layers,  and  is  on  this  account  extensively  used  in  the 
manufacture  of  paint  and  linoleum.  Other  oils,  such  as 
cotton  seed  and  rape  seed  oils  are  often  blown  with  air 
at  about  150°  in  order  to  increase  their  viscosity  for 
use  as  lubricants  when  mixed  with  mineral  oils. 

The  diminution  of  the  iodine  values  of  cotton  seed  or 
rape  seed  oils  on  treatment  with  a  current  of  air  at  about 
120°,  may  easily  be  demonstrated  by  a  small  laboratory 
experiment. 

Examples  of  the  application  of  the  results  of  iodine 
value  determinations  in  the  identification  of  the  oils  and 
fats,  and  the  qualitative  and  approximate  quantitative 
analysis  of  mixtures,  will  be  found  below.  (See  pp.  109 
et  seq.) 

(b)  Saponification  Value. — The  saponification  value 
expresses  the  number  of  milligrams  of  potassium 
hydroxide  required  for  the  complete  saponification  of 
i  gram  of  the  fat.  It  is  thus,  in  the  case  of  pure  fats, 
inversely  proportional  to  the  mean  molecular  weight  of 
the  acid  radicles  present,  more  alkali  being  required  to 
saponify  a  given  weight  of  fat  consisting  of  acid  radicles 
of  low  molecular  weight  combined  with  the  glycerol 
residue  than  an  equal  weight  of  fat  containing  acid 
radicles  of  higher  molecular  weight.  On  this  account, 
the  saponification  value  affords  a  means  of  identifying 
or  detecting  certain  fatty  oils  and  fats,  and  at  the  same 
time,  of  detecting  admixed  unsaponifiable  matter,  such 
as  paraffin,  the  presence  of  which  will  obviously  tend  to 
give  an  abnorally  low  value. 

From  the  following  description  of  the  process  for 
determining  the  saponification  value  it  will  be  seen  that 
some  of  the  potash  used  in  the  determination  will  go  to 
neutralise  any  free  fatty  acids  which  the  fat  may  contain. 
If  the  amount  of  free  fatty  acids  should  be  considerable 


96        Industrial  Organic  Analysis 

(see  below,  under  "  Acid  Value  "),  then  the  saponifica- 
tion  value  found  should  be  corrected  to  give  the  true 
"  Ester  Value  "  of  the  fat,  by  subtracting  the  weight  of 
free  fatty  acids  from  that  of  the  fat  taken  and  the  weight 
of  potash  required  for  the  neutralisation  of  the  acids 
from  the  weight  of  the  potash  used  up  in  the  determina- 
tion. Similarly,  in  the  case  of  notable  quantities  of 
added  unsaponifiable  matter  being  present,  the  weight 
of  the  latter,  if  estimated,  may  be  subtracted  from  the 
weight  of  the  sample  used  in  the  determination  in  order 
to  arrive  at  the  saponification  value  of  the  fat  present 
in  the  mixture.  The  actual  details  of  these  calculations 
will  be  obvious  from  what  is  said  below  on  the  calcula- 
tion of  the  saponification  and  acid  values. 

For  the  determination  of  the  saponification  value  the 
following  solutions  will  be  required  :— 

Alcoholic  Caustic  Potash  Solution. — 35  to  40  grams  of 
potassium  hydroxide,  purified  by  alcohol,  are  dissolved 
in  about  40  c.c.  of  water,  and  made  up  to  1,000  c.c.  with 
96  per  cent,  alcohol ;  the  latter  should  be  tested  before 
use  by  boiling  a  few  c.c.  with  an  equal  bulk  of  con- 
centrated caustic  alkali  solution,  whereupon  only  a 
very  faint  yellow  coloration  should  be  produced.  The 
admixture  of  the  aqueous  potash  and  the  alcohol  may 
be  facilitated  by  shaking  and  warming.  After  standing 
for  a  day,  the  clear  liquor  is  decanted  or  filtered  from 
any  sediment  which  may  have  been  deposited,  and 
preserved  in  a  bottle  furnished  with  a  well-fitting 
rubber  stopper,  so  that  it  will  be  kept  out  of  contact 
with  atmospheric  carbon  dioxide.  The  solution  should 
only  become  light  yellow  on  prolonged  standing  if  the 
alcohol  used  in  its  preparation  was  sufficiently  pure; 
otherwise  it  may  turn  dark  brown. 

Standard  Hydrochloric  Acid  of  half  Normal  Strength. — 
In  the  determination  1-5  to  2  grams  of  the  fat  are  weighed 
into  a,  200  c.c.  Jena  conical  flask,  and  25  c.c.  of  the 


The  Fatty  Oils  and  Fats          97 

alcoholic  potash  solution  are  added  from  a  pipette.  At 
the  same  time,  a  blank  test  is  started,  25  c.c.  of  the  same 
solution  being  measured  into  a  similar  flask,  from  the 
same  pipette,  in  exactly  the  same  way  ;  this  portion 
is  to  be  treated  in  exactly  the  same  way  as  the  portions 
containing  fat,  in  order  that  errors  owing  to  absorption 
of  carbon  dioxide,  and  other  causes,  may  be  eliminated. 
The  flasks  are  fitted  with  simple  tube  condensers,  about 
12  inches  long,  by  means  of  rubber  stoppers,  and  the 
contents  are  kept  gently  boiling  on  a  water  bath,  and 
carefully  agitated  from  time  to  time  in  order  to  hasten 
the  saponification.  This  part  of  the  process  will  be 
complete  when  a  clear,  homogeneous  liquid,  free  from 
particles  of  fat,  has  been  obtained,  the  time  usually  taken 
being  from  20  minutes  to  half  an  hour  ;  if,  however,  the 
presence  of  appreciable  amounts  of  unsaponifiable 
matter  is  suspected,  the  boiling  may  be  prolonged 
somewhat,  the  mixture  being  frequently  agitated  in 
order  to  prevent  the  occlusion  of  unsaponified  fat. 
The  amount  of  potash  left  over  from  each  of  the 
saponifications,  as  well  as  that  in  the  blank  test,  is  now 
estimated  by  titrating  the  hot  solutions  with  semi- 
normal  hydrochloric  acid,  adding  i  c.c.  of  a  i  per  cent, 
alcoholic  solution  of  phenol  phthalein  in  each  case  ; 
if  the  saponification  products  should  have  set  to  a  jelly, 
owing  to  loss  of  alcohol,  a  sufficient  quantity  of  warm 
alcohol,  previously  neutralised  towards  phenol  phthalein, 
to  give  a  clear  solution,  should  be  added. 

From  the  difference  in  the  titers  in  the  blank  test  and 
in  the  actual  test  the  number  of  milligrams  of  potash 
required  to  saponify  i  gram  of  the  sample  may  be 
arrived  at  by  a  simple  calculation. 

Examples  of  the  application  of  the  results  of  saponifica- 
tion value  determinations  in  the  identification  of  oils  and 

I.O.A.  7 


98        Industrial  Organic  Analysis 

fats,  and  the  analysis  of  mixtures,  will  be  given  below. 
(See  pp.  109  et  seq.) 

Besides  furnishing  a  useful  constant  for  the  identifica- 
tion or  detection  of  certain  oils  and  fats,  the  saponifica- 
tion  value  determination  often  admits  of  the  detection 
of  unsaponifiable  matter,  which,  if  present  in  appreciable 
quantity,  will  be  seen  as  an  immiscible  liquid  or  insoluble 
solid  on  diluting  the  neutralised  saponification  product 
with  distilled  water  ;  any  soap  which  may  separate  out 
at  this  stage  will  redissolve  on  warming.  If  it  is  desired 
to  examine  the  unsaponifiable  matter,  however,  the 
saponification  must  be  repeated  on  a  larger  scale.  (See 
below,  under  "Unsaponifiable  Matter.") 

(c)  The  Reichert-Wollny  Value.— The  Reichert-Wollny 
value  expresses  the  number  of  cubic  centimetres  of 
decinprmal  sodium  hydroxide  solution  required  to 
neutralise  the  soluble  volatile  acids  obtained  from 
5  grams  of  the  fat  by  the  Reichert  distillation  process  as 
modified  by  Wollny.  This  method  of  analysis  does  not 
yield  absolute  results,  as  do  the  iodine  and  saponification 
value  determinations,  as  the  whole  of  the  soluble  volatile 
acids  is  not  necessarily  estimated,  but  only  that  portion 
of  them  which  will  be  obtained  when  the  process  of 
distillation,  etc.,  is  carried  out  according  to  certain 
specified  directions.  The  Reichert-Wollny  process, 
which  is  described  below,  is  essentially  the  same  as 
the  original  Reichert-Meissl  process,  the  only  difference 
being  the  greater  precision  which  Wollny  gave  to  the 
method  in  prescribing  the  exact  dimensions  of  the 
apparatus  to  be  used  and  other  practical  details. 

The  first  stage,  consisting  in  the  saponification  of  the 
fat,  may  conveniently  be  carried  out  by,  the  method  of 
Leffman  and  Beam  :  5  grams  of  the  filtered  fat  are 
placed  in  a  flask  and  mixed  with  20  c.c.  of  a  glycerol- 
soda  solution,  prepared  by  dissolving  100  grams  of  pure 
sodium  hydroxide,  free  from  carbonate,  in  100  c.c.  of 


The  Fatty  Oils  and  Fats          99 

water,  allowing  to  stand  till  clear,  and  then  mixing 
20  c.c.  thereof  with  180  c.c.  of  pure  glycerol.  The 
mixture  is  heated  over  a  naked  flame  with  continual 
shaking,  till  it  has  become  perfectly  clear  and  homo- 
geneous, the  process  requiring  about  5  minutes.  100  c.c. 
of  water  which  has  been  boiled  for  at  least  10  minutes 
are  added,  and  the  soap  is  dissolved  by  heating.  40  c.c. 
of  normal  sulphuric  acid  and  about  I  gram  of  pumice 
in  small  lumps  are  added,  and  the  flask  is  at  once  con- 


FIG.  8. — Reichert-Wollny  Distillation  Apparatus. 

nected  up  to  the  bulb  tube  and  condenser,  for  the 
distillation.  The  dimensions  and  form  of  the  flask,  bulb 
tube,  condenser  and  receiver  are  shown  in  the  accom- 
panying illustration.  The  complete  apparatus  may  be 
obtained  from  any  dealer  in  chemical  laboratory 
appliances.  The  flask,  which  rests  on  a  piece  of  asbestos, 
having  a  circular  hole  in  the  centre,  5  cm.  in  diameter,  is 
now  very  gently  heated  until  the  insoluble  fatty  acids 
have  been  completely  melted,  after  which  the  heat  is 
increased  so  that  no  c.c.  are  distilled  over  in  the  course 
of  30  minutes,  as  nearly  as  possible.  The  distillate  is 

7—2 


ioo      Industrial  Organic  Analysis 

shaken,  filtered  into  a  ioo  c.c.  measuring  flask  and  ioo  c.c. 
of  the  filtrate  are  titrated  with  decinormal  baryta  or 
soda  solution,  using  half  a  c.c.  of  a  i  per  cent,  alcoholic 
solution  of  phenol  phthalein  as  indicator.  A  blank  test 
should  also  be  made  :  the  difference  between  the  number 
of  cubic  centimetres  of  decinormal  alkali  used  in  the 
actual  test  and  the  blank  test  is  the  Reichert-Wollny 
number. 

By  the  above  process  such  fats  as  yield  an  excep- 
tionally large  proportion  of  volatile  acids  may  be  detected 
and  approximately  estimated  in  mixtures  with  other 
fats.  The  majority  of  fats  yield  Reichert-Meissl  or 
Reichert-Wollny  values  below  i-o.  Of  the  fats  included 
in  the  table  on  p.  106,  only  the  following  three  yield 
higher  values  :— 

Reichert-Meissl 

Value. 

'  Cocoar\ut  oil  . .          . .          , .  7  to  8-4 

Palm  kernel  oil         . .          . .  5  ,,  6-5 

Butter  fat 26  „  33 

Other  fats  exist  which  give  high  values,  but  these 
will  not  be  considered  here.  It  will  be  noticed  that 
Reichert-Meissl  values  are  most  frequently  given  in  the 
literature  ;  these,  if  carefully  determined,  would,  how- 
ever, not  differ  appreciably  from  the  numbers  which 
would  be  obtained  by  the  Reichert-Wollny  process. 

As  would  be  expected,  the  fats  which  yield  a  large 
proportion  of  volatile  acids  have  correspondingly  high 
saponification  values,  owing,  of  course,  to  the  presence 
of  large  proportions  of  acid  radicles  of  low  molecular 
weight.  Thus,  butter  fat  yields  notable  quantities  of 
butyric  acid  and  the  acids  with  even  numbers  of  carbon 
atoms  in  the  molecule,  immediately  following  the  latter, 
in  going  up  the  series  of  the  fatty  acids,  while  cocoanut 
and  palm  kernel  oils  are  characterised  by  the  unusually 
large  amount  of  lauric  acid,  C12H24O2,  and  myristic  acid, 
CuHasOa,  which  may  be  obtained  from  them. 

The  Reichert-Wollny  process  has  been  adopted  as  the 


The  Fatty  Oils  and  Fats        101 

standard  method  in  Great   Britain   for  the  analysis  of 
butter  and  margarine. 

Examples  of  i-ts  application  will  be  given  below. 

(d)  Acid  Value. — The  acid  value  is  the  number  of 
milligrams  of  potassium  hydroxide  required  to  neutralise 
the  free  fatty  acids  contained  in  one  gram  of  the  oil  or  fat. 

For  the  determination  a  convenient  quantity,  say, 
25  grams  of  the  oil  or  fat,  are  weighed  off  in  a  flask  and 
treated  with  50  c.c.  of  metltylated  spirit  which  has 
previously  been  neutralised  with  sodium  hydroxide 
solution,  using  phenol  phthalein  as  indicator.  Solid  fats 
must  be  melted  so  that  they  can  be  intimately  mixed 
with  the  spirit.  The  mixture  is  kept  well  shaken  while 
it  is  quickly  titrated  with  decinormal  sodium  hydroxide 
solution,  in  presence  of  phenol  phthalein  as  indicator. 
On  standing  for  a  short  while  the  pink  colour  indicating 
the  end  point  of  the  titration  will  disappear,  partly 
owing  to  absorption  of  carbon  dioxide  from  the  air,  and 
partly  owing  to  the  saponification  of  the  fat  ;  the 
titration  should,  of  course,  not  be  continued  on  this 
account.  If  25  grams  of  fat  have  been  used,  then  the 
number  of  cubic  centimetres  of  decinormal  soda  used, 
multiplied  by  0-224,  gives  the  acid  value.  It  is  some- 
times the  custom  to  calculate  the  percentage  of  free  fatty 
acids  expressed  as  oleic  acid  ;  this  figure  may  be  obtained 
by  multiplying  the  acid  value  by  0-53.  If  it  should  be 
necessary  to  correct  the  saponification  value  to  give  the 
ester  value  of  the  fat,  eliminating  the  effect  of  the  free 
fatty  acids  (see  p.  96),  then  the  weight  of  free  fatty  acids, 
expressed  as  oleic  acid,  in  the  sample  taken  for  the 
saponification  value  determination  may  be  calculated 
and  deducted  in  order  to  find  the  weight  of  fat  actually 
taken,  and  the  weight  of  potash  required  to  neutralise 
the  acids  present  deducted  from  the  total  amount  of 


IO2      Industrial  Organic  Analysis 

potash  required  to  saponify  the  sample ;  the  correction 
will,  however,  be  unnecessary,  unless  the  percentage  of 
fatty  acids  be  large.  In  the  case  of  cocoanut  oil,  the  free 
fatty  acids  are  sometimes  calculated  to  lauric  acid,  to 
find  the  percentage  of  which  the  acid  value  is  multiplied 
by  0-357.  The  Kottstorfer  value  expresses  the  number 
of  cubic  centimetres  of  normal  sodium  hydroxide  solution 
required  to  neutralise  the  free  fatty  acids  in  100  grams  of 
fat. 

As  mentioned  above,  crude  vegetable  oils  and  fats 
almost  invariably  contain  appreciable  quantities  of  free 
fatty  acids  ;  thus,  a  sample  of  crude  cocoanut  oil  will 
probably  be  found  to  have  an  acid  value  lying  somewhere 
between  5  and  20,  while  a  sample  of  the  same  fat  which 
has  been  refined  for  edible  purposes  will  show  an  acid 
value  of  less  than  0-3.  Oils  and  fats  which  are  to  be  used 
as  lubricants  must  not  contain  large  proportions  of  free 
fatty  'acids,  as  these  would  act  on  the  bearings  to  form 
metallic  soaps  which  would  exert  a  clogging  effect. 
The  acid  values  of  such  oils  and  fats  should  in  any  case 
lie  well  under  3,  while  in  many  cases  it  may  be  well  to 
insist  on  an  acid  value  not  exceeding  I. 

(e)  Separation  and  Examination  of  the  Unsaponifiable 
Matter. — Under  this  heading  will  be  described  the  method 
for  isolating  and  examining  the  unsaponifiable  matter 
which  is  a  normal  constituent  of  the  fatty  oils  and  fats. 
The  detection  and  estimation  of  foreign  unsaponifiable 
matter,  such  as  paraffin  wax,  will  be  treated  of  later. 
The  determination  of  the  amount  of  naturally  occurring 
unsaponifiable  matter  is  of  no  importance  as  a  method 
of  characterising  the  oils  and  fats,  owing  to  the  varying 
results  obtained  with  different  samples  of  the  same 
product.  Further,  the  amount  of  this  constituent 
present  is  usually  very  small,  generally  of  the  order  of 
about  i  per  cent.,  though  in  some  cases,  as  for  example 
in  certain  marine  animal  oils  and  exotic  vegetable  fats, 
it  may  amount  to  considerably  more.  It  may  also  be 
mentioned  that  the  waxes,  which  consist  of  esters  of 
higher  insoluble  alcohols,  yield  from  35  to  55  per  cent,  of 


The  Fatty  Oils  and  Fats        103 

"  unsaponifiable  matter,"  owing  to  the  fact  that  these 
alcohols,  which  take  the  place  of  the  soluble  glycerol  in 
the  oils  and  fats,  are  obtained  together  with  the  unsaponi- 
fiable matter  proper,  which  in  this  case  consists  mainly 
of  higher  hydrocarbons. 

It  has  already  been  pointed  out  that  the  qualitative 
examination  of  the  naturally  occurring  unsaponifiable 
matter  of  the  commoner  fatty  oils  and  fats  affords  a 
valuable  means  of  characterising  the  vegetable  products 
on  the  one  hand,  and  the  animal  products  on  the  other  ; 
it  is  of  especial  value  in  the  detection  of  the  former  in 
admixture  with  the  latter.  In  what  follows  it  is  assumed 
that  the  fat  is  free  from  paraffin  wax  or  other  similar 
unsaponifiable  matter,  the  presence  of  which  would 
vitiate  the  results  obtained  by  the  operations  to  be 
described  ;  special  methods  for  dealing  with  such  con- 
tingencies are,  however,  described  in  some  of  the  works 
mentioned  at  the  end  of  this  chapter.  (See  "  The 
Chemical  Technology  and  Analysis  of  Oils,  Fats  and 
Waxes/'  by  Lewkowitsch,  and  "  Fatty  Foods,"  by 
Bolton  and  Re  vis.) 

The  isolation  of  the  unsaponifiable  matter  may  be 
carried  out  as  follows  :  50  grams  of  the  fat  are  saponified 
as  described  under  the  heading  "  Titer  Test  "  (p.  89), 
using  proportionate  amounts  of  alkali,  water  and 
alcohol.  The  residual  soap,  obtained  after  the  evapora- 
tion is  complete,  is  dissolved  in  300  c.c.  of  hot  water  and 
transferred  to  a  separating  funnel  of  about  700  c.c. 
capacity,  using  about  50  c.c.  of  water  to  rinse  out  the 
dish.  After  cooling,  the  soap  solution  is  extracted  with 
three  successive  portions  of  about  100  c.c.  of  ether.  In 
order  to  avoid  the  formation  of  emulsions  which  only 
clear  after  prolonged  standing,  violent  shaking  should 
be  avoided,  the  mixture  being  agitated  with  a  rotary 
motion.  In  any  case,  it  will  probably  be  necessary  to 
allow  the  mixture  to  stand  overnight  after  the  first 
treatment  with  ether ;  the  addition  of  a  little  alcohol 


104      Industrial  Organic  Analysis 

may  often  accelerate  the  breaking  down  of  an  emulsion. 
The  ethereal  extracts  are  united  and  washed  with  water 
in  order  to  remove  small  quantities  of  dissolved  soap, 
evaporated  to  dryness,  dried  in  a  steam  oven  and,  if 
desired,  estimated  by  weighing. 

The  identification  of  the  alcohols  cholesterol  and 
phytosterol,  which  in  the  majority  of  cases  constitute 
the  bulk  of  the  unsaponifiable  matter  obtained  from 
animal  and  vegetable  fats  respectively,  is  carried  out  by 
microscopic  examination  of  their  crystalline  structure, 
and  the  determination  of  the  melting  points  of  their 
acetyl  derivatives.  The  first  mentioned  method  is  not 
described  here,  as  the  indications  afforded  by  it  are  Jess 
certain  than  those  afforded  by  the  method  of  deter- 
mining the  melting  points  of  the  acetyl  derivatives,  or,  as 
it  is  called,  the  phytosteryl  acetate  test,  the  appearance 
of  the  crystals  varying  according  to  the  conditions  of 
crystallisation,  especially  when  both  alcohols  are  present. 
The  method  of  microscopic  examination  which  may, 
with  practice,  furnish  useful  indications,  which  may  be 
confirmed  by  the  phytosteryl  acetate  test,  will  be  found 
described  in  some  of  the  larger  works  mentioned  at  the 
end  of  this  chapter. 

Phytosteryl  Acetate  Test. — This  test  depends  on  the  fact 
that  the  melting  point  of  phytosteryl  acetate  lies  some 
10°  above  that  of  cholesteryl  acetate,  and  that  the 
melting  point  of  the  latter  is  raised  by  the  presence  of 
the  former.  It  is,  therefore,  possible  to  apply  the  test 
to  the  detection  of  the  adulteration  of  animal  fats  such  as 
butter  fat  or  lard,  with  vegetable  fats,  as,  for  example, 
cotton  seed  oil,  cocoanut  oil,  etc. 

The  unsaponifiable  matter,  obtained  as  described 
above,  is  boiled  with  a  small  quantity  of  acetic  anhydride 
in  a  small  porcelain  dish  covered  with  a  watch  glass,  for 
a  minute  ;  the  excess  of  acetic  anhydride  is  then  evapo- 
rated off  and  the  residue  is  crystallised  at  least  three 
times  from  a  few  cubic  centimetres  of  absolute  alcohol, 


The  Fatty  Oils  and  Fats        105 

washing  the  separated  crops  with  small  quantities  of 
95  per  cent,  alcohol.  The  melting  point  of  cholesteryl 
acetate  is  114-3°  to  114-8°,  and  that  of  phytosteryl 
acetate  is  125°  to  130°.  If  only  one  of  these  substances 
is  present  either  of  the  above  melting  points  will  readily 
be  obtained  ;  if  both  are  present,  the  melting  point  of 
the  mixture  will  lie  somewhere  between  115-4°  an<^  I25°« 
In  dealing  with  the  unsaponifiable  matter  from  animal 
fats  supposed  to  be  adulterated  with  vegetable  fats,  it 
is  usual  to  repeat  the  crystallisation  some  six  or  seven 
times  ;  the  successive  crops  will  then  contain  increasing 
quantities  of  the  less  soluble  phytosteryl  acetate,  if  this 
be  present,  and  the  melting  point  of  the  material  will  be 
raised  on  each  successive  crystallisation.  If,  on  the 
other  hand,  no  phytosteryl  acetate  is  present,  the 
melting  point  will  not  rise  above  116°  C.  on  repeated 
crystallisation. 

THE  IDENTIFICATION  OF  FATTY  OILS  AND  FATS,  AND 
THE  ANALYSIS  OF  FATTY  MIXTURES. 

Examples  of  the  application  of  the  analytical  methods 
described  above,  together  with  special  tests  by  which 
certain  oils  and  fats  may  be  recognised  or  estimated, 
will  be  given  under  this  heading.  The  analysis  of  fatty 
mixtures  may  in  many  cases  be  a  matter  of  extreme 
difficulty,  requiring  considerable  experience  in  this 
branch  of  analytical  chemistry  ;  the  examples  given 
below  will  only  include  comparatively  simple  cases  of  the 
detection  of  adulterations  which  have  been  known  to 
occur  in  actual  practice  ;  only  a  limited  number  of  the 
fatty  oils  and  fats  will  be  dealt  with,  while  the  analytical 
processes  described  in  this  chapter  are  by  no  means 
completely  representative  of  the  means  available  to  the 
expert.  In  the  table  on  p.  106  will  be  found  a  list  of 
some  of  the  more  important  fatty  oils  and  fats,  together 
with  their  physical  and  chemical  constants  which  may 
be  determined  by  methods  already  described. 


106      Industrial  Organic  Analysis 


1111    Ml          ii    1 

^     11     2 

« 

3     -a  2      «  ti 

S3 

•£    .   «       §      ^       §*5  "§                ^-t-r       rt 

£     5         ^S 

OO.^             ^               Q-              Qr^nrt               Q^             SSCnOt/1 

5         ~M        ^     . 

1 

c  a       T  ** 

1 

£ 

=•.5!  is    fjfjS    |;-i. 
11  f  1  1  flflf  Illjf 

if     I'^ll 

PLiCAlfePnP-ll-lfefe                   r-*. 

CQ       S           H» 

||g 

O           O\         O           M                        N         00 

&  1  1  1  1  8  r  i  s 

1O           CO                 fx 

T    T     T 

H          d               -r 

S^       o 

W               C«                       M 

fit:- 

vi         u^        vo                     >£          0         <0 
vpcOOIMioHcON                Q 

O                                                      fO                                                           « 

5o     S5     2-     ^     t     ^     - 

M                 p» 

en        < 

H 

O 

*,0 

str 

1       1       1       1       1       1       1       1          1 

1                      1 

rt    V 
0    P 

S-SggS^o        g 

|    S       3! 

1 

rt  o 

[  i  ri  1  1  r  i  ^ 

o*o\ooooo<^cyioo           oo 

i   i     oi 

00           &               00 

M1 

I 

MHI^                                 MOOO                       CO 

CO 

< 

1            £ 

11       i       ~        I        HI          I 

M               H               H                                 H 

H 

1   i 

1 

JO 
O           O           N           CO          rj-          O            O 
OSWWMQO'O                10 

CO          N                0 

~            £ 

II                   ^ 

^ONHOO\OOOO            co 

. 

£ 

!? 

CO 

> 

co           .10      .rv^Jio      •cr)f)^      •  O       •  °^         C_5C 

H 

^5^O^O\o      ON^^O      ^^ON^^         QV 

§\  o     S  ^   O>         ^   O* 

O 

£ 

o^    -i    "iSco^i    ?i    "i    "I         1? 

1   i^^Vl      ?l 

tj 
I 

+j                •             •      *£   O\          <>  *^  P1          CTi                      "*• 

*<?*$    <  ^ 

C/) 

5  . 

^  v  i 

_ 

o£ 

••g       |        °       ^-      .-g       ^       rg                    ,_.' 

c 

1s 

is^s1!***     o 

1    a    A   1    i    I   1    1       « 
a  •  3  •  $  a   1  £   5   S     o 

1  1   1 

^                s/)                          ** 

The  Fatty  Oils  and  Fats        107 


SO 

g 

11 

II 

>. 

i 

§ 

1 

c 

<S               ^ 
'U               ^ 

I     i! 

Medicine,  cu 

0 

1 

Lubricant  fo 
machinery, 
treatment. 

Chocolate,  c 
tionery,  r 

1 

1 

1 

'oj            "oj                                               8 
"O*             •O"                                     C8*             "c 

0            o                             °            rt 

§               §                    1              1"          8 
C^             -W                   fa              fe              fe 

«   .       aV 
o|       ?S 

sl       e^ 

ll    «I§ 

^      h 

c, 

0 

00 

10                         O                   tx                 OO 

oo 

I 

I 

I 

J^ 

8        f            III 

NT)-                     OO                  CO               O 

i 

Tt- 

* 

C-l                          CO                  TJ-                  -n- 

5 

1 

! 

| 

I 

1      1         1       1       1 

O                       C-l                               CO                     CO                     M 

1       1 

CO 

cJi               M                     *>•*•+ 

to 

1 

1 

CO 

CO 

1 

CO 

t-N 

I 

10 

01           •_                co            o>           oo 

1!     I    1    r 

N                ^                     M                 TJ-               CO 

I  ' 

00 

ON 
Oi 

M 

Oi 

i" 

O                O                      CO               o                10 

10           <o                co            5            01 

o\ 

5 

CO 

5 

I 

1 

Oi 

i      i        I      I      i 

k   " 

R 

Oi 

CO 

1 

1 

CO 

1 

CO 
CO 

O                 Oi                     CO               .CO              OO 

IT        i            ••* 

00 

1        1 

* 

ro 

10 

M                   ^                       OO                   •*                  CO 

0 

CO 

OO 

<o 

CO 

oo 

R 

rf 

10 

1 

r^            o                 o           oo            •<*• 

MM                          10                  rf                  •* 

C                00 

I 

CO 
Ol 

J, 

1O 

co            ol                  J,             J,              I 

I,    1 

*8 

dj 

o-z     \^_    o^^o^    o^ 

0~ 

f^ 

*o 

c<>  M 

o 

ffi      "                  C)   Q     "      ^    ffi      IJ"      g,    ffi      II                  W      II     £ 

EC     ^ 

0 

0 

<-»& 

—    II 

^ 

a!ln^^> 

^CJ  oo   CJ 

3.o 

1 

10  1 

lr 

1    M  M  ^ 

o  o   tx  ^  ^-o  *i   ^^      W^  (2          1 

M 

On" 

/-_, 

1    rt  II  °° 

o     »ooo     ^               in  o    vo   o     "^oo    ^  lOoc 

°O  u^  Ol    M   P* 

< 

o 

^£ 

OO    Q 

*'"'       ?%£  °^  "-'    2J?« 

1     O    M  OO      ^j 

< 

<£ 

<«         <0i              <               <«         <« 

^ 

•5 

j| 

• 

1-1                                        o 

•              • 

O 

= 

1 

1 

1 

S 

0 

S3 

I  .'  .  i     !   i    • 

a       S         III 

'5 

,J 

* 

0 

(2 

"rt              o                  3              ®             5 
Pi            O                PQ            P3            S 

5     3 

io8      Industrial  Organic  Analysis 

In  the  first  place,  the  determination  of  the  acid  value 
and  the  saponification  value  or  the  unsaponifiable  matter 
will  show  how  far  the  sample  consists  of  neutral  glycerides. 
The  specific  gravity,  which  will  generally  only  be  deter- 
mined in  the  case  of  the  liquid  oils,  will  not  afford  much 
definite  information  ;  in  some  cases  abnormal  values 
may  give  useful  indications  of  adulterants  to  be  looked 
for,  or  confirmation  of  conclusions  arrived  at  by  other 
means.  The  determination  of  the  iodine  value,  on  the 
other  hand,  is  of  great  use  in  the  analysis  of  mixtures, 
owing  to  the  considerable  variations  in  this  constant 
with  the  different  oils  and  fats.  As  regards  the  saponi- 
fication value,  it  will  be  noticed  that  the  majority  of  the 
oils  and  fats  have  values  lying  in  the  neighbourhood  of 
190  ;  rape,  castor,  and  cod  liver  oils  will  be  seen  to  be 
characterised  by  lower,  and  butter  fat  and  cocoanut  and 
palm  kernel  oils  by  higher,  values.  For  reasons  already 
pointed  out,  samples  having  high  saponification  values 
should  be  examined  by  the  Reichert-Wollny  process  for 
volatile  acids.  The  limitations  of  the  use  of  the  deter- 
mination of  the  melting  points  of  oils  and  fats  for  their 
identification,  has  already  been  pointed  out.  Regarding 
the  methods  for  examining  the  insoluble  fatty  acids,  the 
titer  test  may  often  be  of  great  value,  while  the  deter- 
mination of,  the  iodine  values  of  the  fatty  acids  may  be 
resorted  to  when  the  original  sample  contains  large 
quantities  of  added  unsaponifiable  matter  from  which  it 
cannot  be  separated. 

In  addition  to  the  chemical  and  physical  methods,  the 
taste  and  smell,  especially  on  warming,  of  the  sample, 
may  afford  useful  indications.  Even  the  non-expert 
may  in  some  cases  be  guided  by  these  methods,  as  for 
example,  in  the  detection  of  fish  oils,  especially  if  com- 
parison be  made  with  genuine  samples.  Matters  may 
sometimes  be  simplified  for  the  analyst  by  considerations 
of  relative  cost  ;  thus  if  a  sample  purporting  to  consist 
of  a  certain  kind  of  oil  or  fat  appears  not  to  be  genuine, 
it  is  obvious  that  the  adulterant  or  substitute  will  only 
consist  of  a  material  which  is  cheaper  at  the  ruling 
market  prices. 

The  Detection  and  Estimation  of  Adulterants  in  Olive 


The  Fatty  Oils  and  Fats         109 

Oil. — Of  the  possible  adulterants  of  olive  oil,  the  follow- 
ing will  be  considered  :  cotton  seed  oil,  sesame  oil,  lard 
oil,  arachis  oil  and  rape  seed  oil. 

Ex.  i.  Cotton  Seed  Oil  in  Olive  Oil. — On  reference  to 
the  list  of  iodine  values  in  the  table,  it  will  be  noticed 
that  olive  oil  possesses  a  lower  iodine  value  than  any  of 
the  adulterants  named  above,  with  the  exception  of  lard 
oil.  Although,  in  exceptional  cases,  perfectly  genuine 
samples  of  olive  oil  have  been  known  to  give  iodine 
values  above  90,  yet  a  sample  showing  an  iodine  value 
above  the  limits  given  in  the  table  should  be  regarded 
with  suspicion. 

Cotton  seed  oil  may  be  detected  in  presence  of  other 
oils  by  the  Halphen  test,  which  is  carried  out  as  follows  : 
About  2  c.c.  of  the  sample  are  mixed  with  an  equal 
volume  of  amyl  alcohol  in  a  test  tube,  and  about  2  c.c. 
of  a  i  per  cent,  solution  of  sulphur  in  carbon  disulphide 
are  added.  The  tube  is  then  placed  in  a  beaker  of  warm 
water  which  is  gradually  heated  to  boiling.  In  the 
presence  of  cotton  seed  oil,  a  characteristic  orange  red 
coloration  will  be  developed  after  heating  in  the  boiling 
water  for  about  10  minutes.  It  should  be  noted  that 
cotton  seed  oil  which  has  been  heated  above  200°  C.  does 
not  respond  to  the  Halphen  test. 

E.  Gastaldi  (see  Journ.  Chem.  Soc.,  Abstracts  ii., 
1912,  p.  1108)  has  found  that  cotton  seed  oil  does  not 
respond  to  the  Halphen  test  when  this  is  carried  out 
with  carefully  purified  amyl  alcohol,  in  vessels  free  from 
all  traces  of  alkali.  Concluding  that  the  reaction 
depends  on  the  presence  of  traces  of  basic  substances, 
this  author  has  modified  the  Halphen  test  as  follows  : 
2  c.c.  of  the  oil  are  treated  with  I  drop  of  pyridine  and 
about  4  c.c.  of  a  i  per  cent,  solution  of  sulphur  in  carbon 
disulphide,  and  the  mixture  is  heated  in  the  water  bath. 


no      Industrial  Organic  Analysis 

This  test  is  more  delicate  than  the  original  Halphen 
test,  and  can  be  recommended  in  its  place.  It  is,  how- 
ever, subject  to  the  same  limitations  as  the  latter.  (See 
above.) 

Other  colour  reactions  for  cotton  seed  oil  are  known, 
but  these  are,  generally  speaking,  not  so  reliable  as  the 
Halphen  test. 

If  a  positive  reaction  for  cotton  seed  oil  has  been 
obtained,  the  extent  of  the  adulteration  may  be  approxi- 
mately calculated  from  the  iodine  value  of  the  sample 
as  follows  :  Supposing  the  iodine  value  found  to  be  93. 
Then,  taking  the  average  iodine  values  of  cotton  seed  and 
olive  oils  as  106  and  83  respectively,  let  x  equal  the 
percentage  of  cotton  seed  oil  in  the  sample. 
100  -  x  106 


Whence  x  =  from  40  to  45  per  cent. 

It  will  be  observed  that  the  difference  between  the 
saponification  values  of  the  two  oils  is  not  sufficiently 
great  to  afford  a  means  for  detecting  the  adulterant  ; 
this  would  also  be  the  case  if  any  of  the  possible 
adulterants  mentioned  above,  with  the  exception  of  rape 
oil,  were  present.  Confirmation  of  the  presence  of 
cotton  seed  oil  might  possibly  be  obtained  from  a  some- 
what high  specific  gravity  and  titer  test,  or  melting  point 
of  the  fatty  acids,  though  the  latter  figures  would  not 
afford  any  basis  for  quantitative  calculations. 

Ex.  2.  Sesame  Oil  in  Olive  Oil.  —  If  the  sample  shows 
a  higher  iodine  value  than  would  be  expected  from 
genuine  olive  oil,  and  gives  a  negative  test  for  cotton 
seed  oil,  then  sesame  oil  should  be  tested  for  by  the 
Baudouin  test  as  follows  :  10  c.c.  of  the  oil  are  treated 
with  10  c.c.  of  concentrated  hydrochloric  acid  and  2  drops 
of  a  2  per  cent,  solution  of  furfuraldehyde  ;  the  mixture 


The  I^atty  Oils  and   Fats        1 1 1 

is  shaken  vigorously  for  one  minute,  and  then  allowed  to 
stand  to  separate;  in  the  presence  of  sesame  oil  the 
aqueous  layer  will  be  coloured  a  strong  crimson. 

This  reaction  allows  of  the  detection  of  small  quanti- 
ties of  sesame  oil  in  mixtures,  and  is  one  of  the  most 
reliable  of  the  colour  reactions  applied  to  oils  and  fats. 
For  these  reasons,  the  addition  of  sesame  oil  to  margarine 
has  been  made  compulsory  in  Germany  and  other 
countries,  in  order  to  provide  a  ready  means  of  recog- 
nising this  article  of  food.  As  the  colouring  matters  in 
margarine  or  butter  sometimes  produce  a  red  coloration 
with  concentrated  hydrochloric  acid,  it  is  necessary, 
when  testing  for  sesame  oil  in  these,  to  wash  the  sample 
two  or  three  times  with  the  concentrated  acid  before 
applying  the  actual  test. 

The  proportion  of  sesame  oil  present  in  olive  oil  may  be 
roughly  calculated  from  the  iodine  value  of  the  sample 
and  the  iodine  values  of  sesame  and  olive  oils,  as  shown 
in  Ex.  i. 

Ex.  3.  Lard  Oil  in  Olive  Oil. — In  this  case  the  adultera- 
tion would  not  be  revealed  by  the  saponification  value, 
while  the  iodine  value  would  only  be  slightly  lowered. 
The  odour  of  lard  oil  would,  however,  be  noticeable  on 
warming,  and  its  presence  could  be  confirmed  by  the 
phytosteryl  acetate  test  (p.  104). 

Ex.  4.  Arachis  Oil  in  Olive  Ofi. — If  the  sample  shows 
a  somewhat  high  iodine  value  and  gives  negative  Halphen 
and  Baudouin  tests,  arachis  oil  should  be  tested  for  by 
Bellier's  test,  and  if  found  to  be  present,  estimated  by 
Renard's  method  as  modified  by  Archbutt.  Both  of 
these  tests,  which  are  described  below,  are  applicable  to 
mixtures  of  liquid  vegetable  oils  and  cocoanut  and  palm 
kernel  oils,  but  break  down  in  the  presence  of  other  solid 
fats,  vegetable  or  animal. 


112       Industrial  Organic  Analysis 

Bellier's  Test  for  Arachis  Oil. — Before  proceeding  with 
the  actual  test  the  following  two  solutions  must  be 
prepared  : — 

Alcoholic  potassium  hydroxide  solution,  made  by 
dissolving  4-25  grams  of  stick  potash,  pure  by  alcohol,  in 
70  per  cent,  alcohol,  and  making  up  to  50  c.c. 

Acetic  acid  solution  of  such  a  strength  that  1-5  c.c. 
will  neutralise  exactly  5  c.c.  of  the  potash  solution. 

To  carry  out  the  test,  I  gram  of  the  oil  is  placed  in  a 
boiling  tube  6  in.  long  and  i  in.  in  diameter,  and  saponi- 
fied by  gently  heating  with  5  c.c.  of  the  alcoholic  potash 
solution,  shaking  well  and  moderating  the  heating  so  as 
to  avoid  loss  of  alcohol  by  evaporation.  When  a  homo- 
geneous clear  solution  has  been  obtained,  1-5  c.c.  of  the 
acetic  acid  are  added,  and  the  tube  is  cooled  in  water  at 
18°  for  at  least  half  an  hour,  with  occasional  shaking. 
50  c.c.  of  70  per  cent  alcohol,  containing  i  c.c.  of  con- 
centrated hydrochloric  acid  per  100  c.c.  are  added,  and 
the  mixture  is  kept  in  water  at  18°  for  an  hour. 

If  5  per  cent,  or  more  of  arachis  oil  was  present  in 
the  original  sample,  a  precipitate,  consisting  chiefly  of 
arachidic  acid  (see  pp.  79,  114),  will  be  formed,  but  if  no 
arachis  oil  was  present,  the  solution  will  remain  clear  or, 
at  the  most,  become  slightly  turbid.  The  presence  of 
solid  animal  or  vegetable  fats,  other  than  cocoanut  or 
palm  kernel  oils,  will  give  rise  to  the  formation  of  a 
precipitate  which  would  mask  the  test  for  arachis  oil. 

Determination  of  Arachis  Oil  as  Arachidic  Acid. 
(Renard's  Method  modified  by  Archbutt.) — The  determina- 
tion is  based  on  the  fact  that,  while  arachis  oil  yields 
some  4  to  5|  per  cent,  of  acids  of  melting  point  74°  to 
75-5°  C.,  consisting  mainly  of  a  mixture  of  arachidic 
acid,  CaoH^Oa,  and  lignoceric  acid,  C24H48O2,  olive  and 
most  other  oils  yield,  at  the  most,  only  traces  of  these 


The  Fatty  Oils  and  Fats        113 

acids.  In  isolating  the  crude  arachidic  acid  advantage 
is  taken  of  the  solubility  of  the  lead  salts  of  the  unsatu- 
rated  fatty  acids,  e.g.,  oleic  acid,  in  ether,  and  the  in- 
solubility of  the  lead  salts  of  the  alphatic  acids  such  as 
stearic  and  arachidic  acids  in  this  solvent.  Having 
removed  the  unsaturated  acids,  the  arachidic  and 
lignoceric  acids  are  separated  from  the  more  soluble 
aliphatic  acids  of  lower  molecular  weight  by  crystallisa- 
tion from  alcohol. 

For  the  determination,  10  grams  of  the  oil  are  saponi- 
fied by  alcoholic  potash,  and  the  soap  which  remains 
after  removal  of  the  alcohol  by  evaporation  is  dissolved 
in  water.     The  soap  solution  is  acidified  with  dilute 
hydrochloric  acid  and  the  free  fatty  acids  are  extracted 
with  ether.     The  ether  is  evaporated  off  and  the  acids 
are  dissolved  in  50  c.c.  of  90  per  cent,  alcohol  by  heating, 
and,  without  allowing  the  solution  to  cool  so  as  to  deposit 
crystals,  5  c.c.  of  a  20  per  cent,  solution  of  lead  acetate 
are  added.     It  will  be  noticed  that  the  amount  of  lead 
acetate  added  is  insufficient  to  precipitate  the  whole  of 
the  fatty  acids  present  as  lead  salts  ;   the  whole  of  the 
arachidic  and  lignoceric  acids  will,  however,  according  to 
Archbutt,  be  contained  in  the  precipitate  ;  by  using  this 
method  of  fractional  precipitation,  the  subsequent  pro- 
cess of  dissolving  out  the  lead  oleate,  etc.,  is  considerably 
shortened.     The  mixture  is  cooled  to  15°,  agitated  and 
allowed  to  stand  for  half  an  hour,   after  which  the 
precipitate  is  filtered  off  and  washed  once  with  ether. 
The  lead  salts  are  then  washed  back  into  the  flask  by 
means  of  ether,  and  digested  with  the  latter  on  the  water 
bath,  brought  back  on  to  the  filter,  the  whole  process 
being  repeated  twice.     The  lead  salts  are  transferred  to 
a  separating  funnel  by  washing  with  ether,  decomposed 
by  dilute  hydrochloric  acid,  and  the  ethereal  layer,  con- 

I.O.A.  8 


H4      Industrial  Organic  Analysis 

taining  the  liberated  acids  in  solution,  is  washed  with 
water  till  free  from  mineral  acid.  The  ethereal  solution 
is  transferred  to  a  flask,  evaporated  to  dryness,  and  the 
solid  acids  thus  obtained  dried  at  100°. 

50  c.c.  of  90  per  cent,  alcohol  are  added  to  the  acids  in 
the  flask,  which  is  then  corked  lightly  and  warmed  until 
the  acids  are  dissolved  ;  the  solution  is  cooled  to  either 
15°  or  20°,  kept  at  this  temperature  for  3  hours,  after 
which  the  crop  of  crystals  is  filtered  off  on  a  Gooch 
crucible  and  washed  with  3  portions  of  10  c.c.  of  90  per 
cent,  alcohol  at  the  same  temperature.  The  filtrate  and 
washings  of  90  per  cent,  alcohol  from  this  and  subsequent 
operations  should  be  measured,  in  order  that  a  correction 
for  the  dissolved  arachidic  acid  may  be  introduced. 
The  washing  should  not  be  carried  out  too  rapidly,  as  the 
alcohol  passing  through  the  filter  should  become  satu- 
rated* with  the  acid.  The  crystals  are  then  washed  with 
70  per  cent,  alcohol ;  these  washings  may  be  rejected,  as 
practically  no  arachidic  acid  will  be  dissolved.  The 
crystals  are  dissolved  from  the  filter  by  warm  ether, 
which  is  evaporated  in  a  tared  flask,  and  the  residue 
dried  and  weighed.  The  melting  point  of  the  acids  is 
determined  in  the  usual  way  ;  if  this  is  found  to  be  under 
71°  the  process  of  recrj^stallisation  and  washing  must  be 
repeated  as  described  above,  the  quantity  of  90  per  cent, 
alcohol  used  being  noted  as  before. 

The  amount  of  arachidic  acid  estimated  to  have  been 
dissolved  by  the  90  per  cent,  alcohol  is  added  to  that 
actually  weighed.  The  amount  of  arachis  oil  in  the 
original  sample  may  then  be  approximately  calculated 
by  multiplying  the  total  arachidic  acid  by  20,  as  it  has 
been  found  that  arachis  oil  yields  about  5  per  cent,  of 
arachidic  acid  melting  at  71  to  72-5  by  the  method 
described. 


The  Fatty  Oils  and  Fats         115 

In  order  to  allow  for  the  dissolved  acid  the  following 
table,  due  to  Archbutt  (abridged),  may  be  used  : — 


Grams  of  Acids  dissolved  by  100  c.c.  of 

Weight  of  Acids 

90%  Alcohol  at 

obtained. 

(grams) 

15°  C. 

17-5°  c. 

20°  C. 

o-i  or  less    . 

0-033 

0-039 

O-O46 

0-3      • 

0-055 

0-064 

0-074 

o-5      • 

0-064 

0-075 

0-085 

0-8  or  more. 

0-070 

O-o8o 

0-091 

In  the  presence  of  solid  animal  and  vegetable  fats,  the 
test  is  not  so  satisfactory  in  its  quantitative  application, 
but  may  still  be  used  qualitatively  in  cases  where 
Bellier's  test  breaks  down  ;  according  to  Lewkowitsch, 
it  is  possible  to  estimate  approximately  arachis  oil  in 
lard  by  Renard's  test.  In  such  cases,  however,  it  may 
be  necessary  to  recrystallise  repeatedly  from  90  per  cent, 
alcohol  before  an  arachidic  acid  melting  over  70°  is 
obtained,  while  a  negative  result  cannot  always  be  taken 
as  evidence  of  the  absence  of  arachis  oil. 

Arachis  oil  is  very  frequently  used  as  an  adulterant 
in  olive  oil ;  as  will  be  seen  from  the  table  on  p.  106, 
the  saponification  value  or  specific  gravity  of  the 
sample  would  not  reveal  the  presence  of  this  adulterant, 
while  the  difference  between  the  iodine  values  of  olive 
and  arachis  oils  is  not  very  great. 

Ex.  5.  Rape  Oil  in  Olive  or  Arachis  Oils. — The 
presence  of  considerable  quantities  of  rape  oil  in  olive  oil 
might  be  indicated  by  a  somewhat  high  iodine  value,  but 

8—2 


ii6      Industrial  Organic  Analysis 

in  the  case  of  arachis  oil  supposed  to  be  adulterated  with 
rape  oil,  no  definite  information  could  be  obtained  from 
the  result  of  the  determination  of  this  constant.  Bear- 
ing in  mind  the  comparatively  low  saponification  value 
of  rape  oil,  it  is  evident  that  samples  of  olive  or  arachis 
oils  adulterated  with  appreciable  quantities  of  rape  oil 
should  show  abnormally  low  saponification  values.  In 
this  connection  it  may  be  mentioned  that  of  the  oils  and 
fats  given  in  the  table  on  p.  106,  which  have  low  saponifi- 
cation values,  rape  oil  is  the  only  one  which  would  be 
likely  to  occur  as  an  adulterant  of  arachis  or  olive  oils. 

The  approximate  proportion  of  rape  oil  present  might 
be  calculated  from  the  saponification  value  of  the  sample 
as  follows  :  Supposing  that  a  suspected  sample  of  arachis 
oil  was  found  to  have  an  iodine  value  of  97,  and  a 
saponification  value  of  187.  Then,  taking  the  mean 
saponification  value  of  arachis  oil  as  193-5,  and  that  of 
rape  oil  as  174*5,  if  x  equals  the  percentage  of  rape  oil 
present, 

100  ~*  193  +  ^l*  =  187, 

100  100 

whence  %  =  from  30  to  35  per  cent. 

In  this  case  the  determination  of  the  specific  gravity 
would  be  of  little  use,  but  a  sample  of  arachis  oil  adulte- 
rated with  rape  oil  might  show  an  abnormally  low 
titer  test. 

As  rape  oil  has  been  shown  to  contain  only  very  small 
amounts  of  arachidic  acid,  a  determination  of  arachis  oil 
as  arachidic  acid  could  be  made  as  described  above,  in 
order  to  check  the  result  obtained  from  the  calculation 
based  on  the  saponification  value. 

Ex.  6.  Cotton  Seed  Oil  in  Rape  Seed  Oil. — The  presence 
of  cotton  seed  oil  as  an  adulterant  in  rape  seed  oil  would 
be  revealed  by  a  higher  saponification  value  than  would 


The  Fatty  Oils  and  Fats        117 

be  expected  from  pure  rape  oil,  and  might  be  confirmed 
by  the  Halphen  test  for  cotton  seed  oil.  The  iodine 
value,  titer  test,  and  specific  gravity  would  all  tend  to  be 
raised  owing  to  the  presence  of  the  adulterant,  though 
these  figures,  excepting,  perhaps,  the  titer  test,  would 
not  be  noticeably  influenced  unless  the  amount  of  cotton 
seed  oil  present  was  large.  The  extent  of  the  adultera- 
tion might  be  approximately  calculated  from  the 
saponification  value  of  the  sample  and  the  mean  saponi- 
fication  values  of  cotton  and  rape  oils,  as  described 
above  in  Ex.  6. 

Ex.  7.  Arachis  Oil,  Sesame  Oil,  or  Cotton  Seed  Oil  in 
Lard. — The  presence  of  any  of  these  oils  might  be  indi- 
cated by  a  high  iodine  value.  Sesame  and  cotton  seed 
oils  could  be  detected  by  the  Baudouin  and  Halphen 
tests  respectively  (see  Exs.  i  and  2),  and  their  presence 
could  be  confirmed  by  the  phytosteryl  acetate  test  for 
detecting  vegetable  oils  and  fats.  Confirmation  by  this 
means  should  always  be  sought  if  cotton  seed  oil  is 
suspected  to  be  present,  as  it  has  been  shown  that  lard 
from  pigs  which  have  been  fed  on  cotton  seed  cakes 
may  itself  give  a  positive  Halphen  test.  If  negative 
Baudouin  and  Halphen  tests  have  been  obtained,  arachis 
oil  should  be  tested  for  and  estimated,  if  present,  by  the 
method  detailed  in  Ex.  4.  Here,  again,  the  phytosteryl 
acetate  test  might  be  applied  to  confirm  the  presence  or 
absence  of  a  vegetable  oil. 

Ex.  8.  Arachis,  Sesame,  or  Cotton  Seed  Oils  in  Cacao 
Butter. — The  adulteration  of  cacao  butter,  or  chocolate 
fat,  with  the  above  mentioned  oils  could  be  detected 
on  the  same  lines  as  indicated  in  the  previous  example, 
excepting,  of  course,  that  the  phytosteryl  acetate  test 
would  in  this  case  not  be  available,  as  cacao  butter  is 
itself  a  vegetable  product.  As  cacao  butter  has  a  fairly 


ii8      Industrial  Organic  Analysis 

low  iodine  value,  the  extent  of  the  adulteration  might 
be  approximately  calculated  as  indicated  in  Ex.  i.  The 
titer  test  and  melting  point  of  the  fatty  acids  of  cacao 
butter  are  fairly  high,  and  would  tend  to  be  lowered 
through  the  presence  of  the  oils  in  question.  A  deter- 
mination of  the  melting  point  of  the  sample  should  also 
be  made,  as  it  is  important  that  fats  used  in  chocolate 
making  should  not  melt  at  too  low  a  temperature. 

Ex.  9.  Cocoanut  or  Palm  Kernel  Oils  in  Lard  or  Cacao 
Butter. — These  adulterants,  if  present  in  appreciable 
quantity,  would  be  indicated  by  a  low  iodine  value  and 
high  saponification  value.  A  fairly  high  Reichert- 
Wollny  value  would  afford  confirmation  of  their  presence. 
Thus,  lard  and  cacao  butter  give  values  of  about  0-5  and 
075,  respectively;  from  the  figures  given  under  the 
description  of  the  method  for  carrying  out  the  Reichert- 
Wollny  determination,  it  might  be  concluded,  for 
example,  that  a  sample  of  lard  showing  a  Reichert- 
Wollny  value  of  about  1-5  would  contain  about  10  per 
cent,  of  cocoanut  oil,  or  rather  less  palm  kernel  oil.  The 
phytosteryl  acetate  test  would  only  be  available  in  the 
case  of  lard.  The  titer  test,  or  melting  point  of  the 
fatty  acids,  would  tend  to  be  lowered  through  the 
presence  of  the  adulterants  under  consideration. 

Ex.  10.  Fish  or  Blubber  Oils  in  Rape  Oil. — Fish  or 
blubber  oils  may  often  be  detected  in  mixtures  by  their 
characteristic  smell  which  becomes  more  noticeable  on 
warming.  The  presence  of  these  adulterants  would 
tend  to  raise  both  the  saponification  and  the  iodine 
values,  especially  the  latter.  Fish  or  marine  animal  oils, 
and  other  drying  oils  such  as  linseed  oil,  may  be  distin- 
guished from,  and  detected  in  the  presence  of  non-drying 
and  semi-drying  oils  such  as  rape  oil,  cotton  seed  oil, 
arachis  oil,  etc.,  by  means  of  the  Hexabromide  test,  which 


The  Fatty  Oils  and  Fats        119 

is  based  on  the  fact  that  the  hexabromo  (or  octobromo) 
derivatives  yielded  by  the  first  mentioned  group  are 
practically  insoluble  in  ether  and  certain  other  solvents  ; 
the  semi  and  non-drying  oils,  on  the  other  hand,  are 
only  capable  of  yielding  lower  bromine  addition  products 
which  are  soluble  under  the  conditions  of  the  test. 
(See  p.  79.) 

Halphen's  modification  of  the  hexabromide  test  is  as 
follows  :  0-5  c.c.  of  the  oil  are  mixed  with  10  c.c.  of  a 
mixture  of  28  parts  by  volume  of  glacial  acetic  acid, 
i  part  of  bromine  and  4  parts  of  nitrobenzene,  in  a  clean, 
dry  test  tube,  the  tube  is  closed  and  shaken,  and  the 
contents  examined.  The  semi-  or  non-drying  oils 
produce,  at  the  most,  only  a  slight  turbidity,  while  the 
fish  or  marine  animal  oils,  or  other  drying  oils,  produce 
a  distinct  precipitate  which  is  not  dissolved  on  the 
addition  of  10  c.c.  of  methylated  ether.  Rape  oil  gives 
a  turbid  liquid  separating  into  two  layers,  but  on  the 
addition  of  10  c.c.  of  ether,  a  clear  liquid  is  formed. 

The  indications  afforded  are,  of  course,  only  of  a 
qualitative  nature,  but  the  test  would  prove  useful  in 
deciding  whether  a  sample  of  rape  oil,  showing  an 
abnormally  high  iodine  value,  had  been  adulterated 
with  a  fish  or  other  drying  oil,  or  with  larger  quantities 
of  a  semi-drying  oil  having  an  iodine  value  higher  than 
rape  oil.  Bolton  and  Revis  state  that  as  little  as  5  per 
cent,  of  fish,  marine  animal  or  drying  oil  may  be  detected 
in  rape  oil,  but  that  the  test  breaks  down  in  the  presence 
of  animal  fats  such  as  beef  tallow  and  lard,  which  pro- 
duce a  precipitate  under  the  conditions  of  the  test. 
Shea  butter  is  stated  to  behave  similarly  to  the  animal 
fats. 

Butter  Adulterated  with  Margarine. — The  detection  of 
margarine  fats  in  butter  fat  may  often  be  a  far  more 


I2O      Industrial  Organic  Analysis 

complex  problem  than  any  of  those  hitherto  considered, 
owing,  firstly,  to  the  fact  that  the  margarine  fat  will 
practically  always  be  a  mixture,  and,  secondly,  to  the 
fact  that  the  chemical  and  physical  constants  of  genuine 
butter  fats  may  vary  within  fairly  wide  limits,  depending 
on  factors  such  as  the  breed  and  age  of  the  cow,  nature 
of  the  fodder,  climatic  conditions  and  general  treatment 
of  the  animals. 

Margarine  usually  differs  from  butter  in  the  com- 
position of  the  fat  only  ;  the  fats  commonly  used  in. 
margarine  making  are  cocoanut  and  palm  kernel  oils, 
beef  tallow  and  lard,  in  addition  to  which  oils  such  as 
cotton  seed,  arachis  and  sesame  oils  are  practically 
always  present.  The  detection  of  quantities  up  to  about 
10  per  cent,  of  margarine  fat  in  butter  fat  may  often  be  a 
matter  of  considerable  difficulty,  especially  as  some  of  the 
constants  of  the  added  fat  mixture  may  approximate 
closely  to  those  of  pure  butter  fat.  For  example,  the 
tendency  for  the  iodine  value  to  be  lowered,  owing  to  the 
presence  of  cocoanut  or  palm  kernel  oils,  might  be 
neutralised  by  the  presence  of  another  oil  with  an  iodine 
value  higher  than  that  of  butter  fat. 

The  most  important  method  for  the  examination  of 
butter  fat  is  the  Reichert-Meissl  or  Reichert-Wollny 
process  for  estimating  the  volatile  acids  yielded  by  the 
fat,  the  latter  having  been  adopted  as  the  standard 
method  for  the  examination  of  butter  fat  in  Great 
Britain.  The  majority  of  butter  fats  yield  Reichert- 
Wollny  values  lying  between  26  and  33,  while  none  of  the 
fats  which  can  be  used  as  adulterants  yield  higher  values. 
A  butter  fat  showing  a  Reichert-Wollny  value  of  24  to  25, 
or  less,  should,  therefore,  be  regarded  with  suspicion, 
though  it  could  not  be  definitely  said  to  be  adulterated 
unless  the  Reichert-Wollny  value  were  below  20,  as 


The  Fatty  Oils  and  F^ats        121 

exceptional  examples  have  been  known  of  perfectly 
genuine  butter  fats  yielding  values  approximating  to  this 
figure.  In  doubtful  cases,  recourse  should  be  had  to 
the  phytosteryl  acetate  test  for  the  detection  of  vegetable 
oils  and  fats.  At  the  same  time  the  colour  tests  for 
sesame  and  cotton  seed  oils  might  be  applied  (see 
Exs.  i  and  2).  As  cocoanut  and  palm  kernel  oils  yield 
moderately  high  Reichert-Wollny  values  (see  p.  100), 
their  presence  would  not  have  so  marked  an  effect  as 
the  presence  of  other  vegetable  oils  or  animal  fats  ;  on 
the  other  hand,  being  vegetable  products,  they  would  be 
revealed  by  the  phytosteryl  acetate  test.  The  presence 
of  cocoanut  or  palm  kernel  oils  only  would  tend  to 
lower  the  iodine  value  and  raise  the  saponification  value, 
while  the  Reichert-Wollny  value  might  not  be  greatly 
affected.  If  the  other  vegetable  oils  mentioned  above 
were  present  as  well,  the  iodine  and  saponification  values 
might  not  be  affected  to  any  noticeable  extent.  Adul- 
teration with  animal  fats  together  with  vegetable  oils 
other  than  cocoanut  and  palm  kernel  oils,  would  produce 
a  greater  effect  on  the  Reichert-Wollny  value  than  in  the 
case  just  considered,  while  the  iodine  value  would 
probably  be  raised,  and  the  saponification  value  lowered. 
The  above  methods  will  suffice,  at  least,  to  distinguish 
between  genuine  butter  and  margarine  or  butter  con- 
taining, say,  25  per  cent,  or  more  of  margarine.  Smaller 
amounts  of  added  fats  may,  in  some  cases,  also  be 
detected,  but  may,  on  the  other  hand,  be  overlooked 
unless  the  sample  is  subjected  to  a  still  more  searching 
examination.  This  is  especially  the  case  if  the  added 
fat  consists  of  cocoanut  oil,  for  the  detection  of  which, 
numerous  special  methods  have  been  proposed.  The 
most  well-known  of  these  is  Polenske's  modification  of 
the  Reichert  distillation  process,  in  which  the  water 


122      Industrial  Organic  Analysis 

soluble  and  insoluble  volatile  fatty  acids  are  separately 
estimated.  Blichfeldt  (Journ.  Soc.  Chem.  Ind.  No.  13, 
Vol.  XXIX.)  has  designed  a  special  distillation  apparatus 
in  which  the  side  tube,  condenser  and  receiver  are  all  in 
one  piece,  and  estimates  the  volatile  fatty  acids  forming 
water  soluble  and  insoluble  silver  salts  ;  this  method 
admits  of  a  sharper  differentiation  between  cocoanut 
and  palm  kernel  oils  and  butter  fat  than  the  methods 
hitherto  mentioned.  Other  methods  for  the  examination 
of  butter  fat  are  the  determination  of  the  refractive 
index  and  the  microscopic  examination  of  the  crystalline 
structure  of  the  fat.  Descriptions  of  these  methods,  as 
well  as  the  Polenske  method,  will  be  found  in  the  works 
of  reference  enumerated  at  the  end  of  this  chapter. 

The  Detection  of  Paraffin  Wax  in  Lard  or  Cacao  Butter. 

—The  following  method  was  devised  by  Polenske  for  the 

detection  and  estimation  of  paraffin  wax  in  lard.     It 

may  also  be  applied  to  the  detection  of  this  adulterant 

in  other  fats. 

100  grams  of  the  sample  are  saponified  by  boiling  in 
a  flask  connected  with  a  reflux  condenser,  with  alcoholic 
potash  in  the  proportions  recommended  above  under  the 
directions  for  the  isolation  of  fatty  acids  from  fats 
(p.  89).  The  paraffin  wax,  if  present,  will  be  noticed  as 
a  layer  floating  on  the  soap  solution.  The  process  of 
saponification  should  be  complete  after  an  hour's  boiling, 
the  mixture  being  well  shaken  from  time  to  time.  If  the 
amount  of  paraffin  wax  present  is  large,  it  may  be 
necessary  to  prolong  the  treatment  in  order  to  secure 
complete  saponification  of  the  fat.  After  removal  of 
the  alcohol  by  evaporation,  the  unsaponifiable  matter, 
including  the  paraffin  wax,  is  separated  as  directed  on 
p.  103.  In  order  to  eliminate  the  naturally  occurring 
unsaponifiable  matter,  and  to  separate  the  paraffin  wax 


The  F7atty  Oils  and  Fats        123 

for  estimation  and  examination,  the  total  unsaponifiable 
matter  is  heated  in  a  test  tube  immersed  in  a  brine  bath 
at  105°  C.,  with  5  c.c.  of  concentrated  sulphuric  acid  for 
half  an  hour.  After  cooling,  the  mixture  is  extracted 
without  previous  dilution,  with  petroleum  ether  which  is 
completely  volatile  below  80°.  The  extraction  having 
been  repeated  twice  with  fresh  quantities  of  petroleum 
ether,  the  petroleum  ether  extracts  are  united  and 
evaporated  to  dryness  in  a  tared  flask.  The  residue, 
consisting  of  paraffin  wax,  is  dried  at  100°  and  weighed. 

From  the  above  examples  it  may  be  seen  how  the 
analytical  processes  described  in  this  chapter  may  be 
applied  in  the  analysis  of  simple  mixtures  of  the  more 
well-known  fatty  oils  and  fats,  or  identifying  unadulte- 
rated samples  of  the  latter.  The  examples  of  adulteration 
given  are  all  comparatively  simple  ;  in  actual  practice, 
however,  problems  may  occur  which  will  require  for 
their  solution  more  extensive  investigations  and  greater 
experience  in  this  branch  of  analytical  chemistry  than 
can  be  obtained  from  the  present  work.  The  student 
wishing  to  pursue  the  subject  further  is  recommended 
to  consult  the  works  mentioned  below,  especially  Lew- 
kowitsch's  "  Chemical  Technology  and  Analysis  of  Oils, 
Fats  and  Waxes." 


REFERENCES. 

"  The  Chemical  Technology  and  Analysis  of  Oils,  Fats  and 
Waxes,"  by  J.  Lewkowitsch,  3  vols.  1909,  Macmillan  &  Co., 
London. 

"  The  Laboratory  Companion  to  Fats  and  Oils  Industries,"  by 
J.  Lewkowitsch.  1901,  Macmillan  &  Co.,  London. 

"  Technologie  der  Fette  und  Oele,"  by  Gustav  Hefter,  3  vols. 
1906  to  1910,  J.  Springer,  Berlin. 

"  Analyse  der  Fette  und  Wachsarten,"  by  Benedikt-Ulzer. 
1908,  J.  Springer,  Berlin. 


124      Industrial  Organic  Analysis 

"  Handbuch  der  Chemie  und  Technologic  der  Oele  und  Fette," 
by  Ubbelohde  and  Goldschmidt,  4  vols.  1908  to  1912,  S.  Hirzel, 
Leipzig. 

"  Allgemeine  und  Physiologische  Chemie  der  Fette,"  by  Ulzer 
and  Klimont.  1906,  J.  Springer,  Berlin. 

"  Laboratoriumsbuch  fur  die  Industrie  der  Oele  und  Fette," 
by  J.  Marcusson.  1911,  W.  Knapp,  Halle. 

"  Untersuchung  der  Mineraloele  und  Fette,"  by  D.  Holde. 
I909,  J-  Springer,  Berlin. 

"  Chemisch-Technische  Untersuchungsmethoden,"  Vol.  III., 
by  G.  Lunge.  1911,  J.  Springer,  Berlin. 

"  Chemisch-Technische  Analyse,"  Vol.  I.,  by  Post.  1908, 
Fr.  Vieweg,  Brunswick. 

"Commercial  Organic  Analysis"  (revised  edition),  Vol.  II.,  by 
Allen.  1910,  J.  and  A.  Churchill,  London. 

"  Fatty  Foods,"  by  Bolton  and  Revis.  1913,  J.  and  A. 
Churchill,  London. 


CHAPTER   IV 

SOAP 

INTRODUCTORY. 

SOAP  consists  essentially  of  the  sodium  or  potassium 
salts  of  the  fatty  or  rosin  acids,  and  is  produced  by  the 
action  of  caustic  alkali  on  fats,  fatty  acids  or  rosin. 
All  soaps  contain  notable  quantities  of  water,  the 
proportion  of  which  may  vary  within  wide  limits, 
while  the  potash  or  soft  soaps  and  some  soda  soaps 
such  as  the  marine  soaps,  contain,  in  addition,  the 
glycerol  which  is  produced  in  the  process  of  saponifying 
the  fat.  Other  possible  constituents  of  soaps  are  men- 
tioned in  the  classification  given  below. 

Soap  is  generally  manufactured  by  treating  fat  with 
a  definite  proportion  of  caustic  alkali  solution,  the  saponi- 
fication  being  effected  either  by  boiling  the  mixture  in 
open  vessels,  heating  under  steam  pressure  in  closed 
vessels,  or,  as  in  the  cold  process,  by  allowing  the  soda 
solution  to  act  on  the  fat  (usually  cocoanut  oil)  at  the 
ordinary  or  slightly  elevated  temperatures.  The  soda 
soaps  produced  according  to  the  two  first  mentioned 
methods  are  separated  from  the  excess  of  water  and 
alkali,  and  the  glycerol,  by  adding  a  solution  of  common 
salt  to  the  boiling  mixture  ;  the  sodium  salts  of  the 
fatty  acids,  being  insoluble  in  sodium  chloride  solution, 
separate  out,  together  with  a  relatively  small  proportion 
of  water,  as  a  molten  layer  on  the  surface  of  the  brine 
which  retains  the  greater  part  of  the  glycerol  and  excess 
of  free  alkali.  Soaps  made  by  the  cold  process,  as  well 
as  the  potash  or  soft  soaps,  are  not  submitted  to  this 
process  of  purification  and,  therefore,  contain  glycerol, 
excess  of  free  alkali  and  a  large  proportion  of  water. 

The  rosin,  or  colophony,  which  is  used  in  conjunction 
with  fats  in  the  manufacture  of  cheaper  soaps,  consists 


126      Industrial  Organic  Analysis 

of  the  residue  which  remains  after  distilling  off  the  oil  of 
turpentine  and  moisture  from  pine  resin  ;  it  is  chiefly 
composed  of  acids  which  dissolve  in  caustic  alkali 
solution  with  the  formation  of  soap-like  products  which 
generally  differ  from  the  true  soaps  produced  from  fats 
in  possessing  greater  alkalinity,  and  being  softer  and 
darker  in  colour.  Rosin  or  colophony  should  be  dis- 
tinguished from  the  rosin  spirit  or  rosin  oils  which  are 
produced  from  it  by  destructive  distillation.  (See 
Chapter  V.,  p.  179.) 

The  fats  most  commonly  used  in  soap  making  are 
tallow,  palm  oil,  recovered  grease,  cocoanut  oil,  cotton 
seed  oil,  maize  oil,  sesame  oil,  hemp  seed  oil,  palm  kernel 
oil,  various  fish  oils,  castor  oil,  olive  oil  and  lard.  In  the 
case  of  edible  fats,  it  is  usually  the  cheaper  grades  which 
are  used  in  soap  making. 

The  cleansing  action  of  soap  depends,  firstly,  on  the 
formation  of  small  quantities  of  free  caustic  alkali  owing 
to  tfre  hydrolytic  action  of  water  on  the  alkali  salts  of 
the  fatty  or  rosin  acids  ;  according  to  the  principles  of 
mass  action,  the  greater  the  proportion  of  water  to  soap, 
the  greater  will  be  the  extent  of  the  hydrolysis,  so  that 
the  concentration  of  free  caustic  alkali  is  to  some  extent 
automatically  regulated  and  kept  low  ;  further,  the 
greater  causticity  of  rosin  soaps  as  compared  with  soaps 
formed  from  fats,  may  be  explained  by  the  fact  that  the 
rosin  acids  are  weaker  acids  than  the  fatty  acids,  and, 
consequently,  their  sodium  salts  are  hydrolysed  to  a 
greater  extent  in  aqueous  solution.  The  removal  of 
dirt  of  a  greasy  nature,  which  cannot  be  effected  by 
means  of  water  alone,  takes  place  as  follows  :  a  small 
portion  of  the  fatty  matter  is  saponified  by  the  caustic 
alkali  and  removed  as  soluble  soap  ;  the  greater  portion, 
however,  is  made  into  an  emulsion  with  the  alkaline 
liquid  and  mechanically  removed  with  the  lather. 

The  various  soaps  on  the  market  may  be  broadly 
classified  as  follows  :— 

(i)  Toilet  Soaps. — These  should  consist  as  far  as 
possible  of  the  neutral  sodium  salts  of  the  fatty  acids 
together  with  moderate  amounts  of  water.  They  should, 
naturally,  contain  no  impurities  or  adulterants,  and, 


Soap  127 

above  all,  no  free  caustic  alkali  or  alkaline  carbonate. 
They  are  often  made  from  lard  or  olive  oil.  The  trans- 
parent toilet  soaps  which  are  produced  by  dissolving 
soap  in  alcohol  and  evaporating  the  clear  solution,  gener- 
ally contain  glycerol,  and  sometimes  also  notable  amounts 
of  sugar. 

(2)  Laundry  Soaps. — These  usually  contain  more  or 
less  free  alkali  as  carbonate  or  hydroxide.     They  are 
usually  made  from  tallow,  palm  oil  or  rosin. 

(3)  Commercial   Soaps. — The   soft   soaps,   which   are 
generally  made  from  fish  oils  or  vegetable  drying  oils, 
contain  glycerol  and  excess  of  potash  and  water.     The 
so-called    hydrated    soaps,    which   are   produced   from 
cocoanut  or  palm  kernel  oils  by  the  cold  saponification 
process,  contain  glycerol  and  excess  of  water  ;   they  are 
also  liable  to  contain  unsaponified  fat  and  free  alkali, 
owing  to  the  incompleteness  of  the  saponification  pro- 
cess.    Marine  soaps  are  made  from  cocoanut  or  palm 
kernel  oils  ;    they  contain  a  large  proportion  of  sodium 
laurate,  (^H^OaNa,  and  sodium  myristate,  CuH27O2Na, 
which  are  more  soluble  in  salt  water  than  the  soaps  from 
other  fats,  and  may,  therefore,  be  used  with  sea  water. 

(4)  Medicated    Soaps. — These    may    contain    phenol, 
cresols,  naphthalene,  and  other  coal  tar  products,  or 
substances  of  a  similar  nature. 

Besides  unsaponified  fat,  free  alkali,  or  free  fatty 
acids,  soaps  may  contain  small  amounts  of  chlorides, 
sulphates,  silicates  and  other  inorganic  impurities 
derived  from  the  materials  used  in  their  manufacture. 
According  to  Allen,  the  following  additions  are  some- 
times made  to  soap  :  oatmeal,  bran,  sawdust,  fuller's 
earth,  chalk,  etc.  Considerable  quantities  of  sand  or 
powdered  quartz  are  used  in  scouring  soaps.  Sodium 
or  potassium  carbonates  are  added  to  scouring  and 
commercial  soaps  in  order  to  increase  their  detergent 
properties  and  also  to  facilitate  lathering  with  hard 
waters.  Iron  compounds,  ochre,  ultramarine  and  other 
colouring  matters  are  added  to  produce  the  effect  of 
mottling.  Sodium  silicate,  aluminate  and  borate, 
petroleum  and  naphtha  may  also  be  found  in  commercial 
soaps. 


128      Industrial  Organic  Analysis 

THE  ANALYSIS  OF  SOAP. 

Among  the  more  important  constituents  to  be  deter- 
mined, are  water,  total  alkali,  fatty  and  rosin  acids,  free 
alkali  as  carbonate  and  hydroxide,  and  combined  alkali 
present  as  soap.  The  estimation  of  adulterants,  or 
fillers  and  legitimate  additions  such  as  alkaline  carbonate 
silicate  and  phenol,  will  also  be  described.  The  following 
scheme  for  the  analysis  of  soap,  which  will'  first  be 
briefly  summarised  here,  is,  in  the  main,  identical  with 
that  due  to  Allen,  Leeds,  and  others,  which  is  adopted 
in  Allen's  "  Commercial  Organic  Analysis/'  1911  edition. 

Scheme  of  Analysis. — The  sample  for  analysis  should, 
in  the  case  of  a  solid  soap,  be  taken  from  the  interior 
of  the  bar  or  cake,  avoiding  the  outer  dried  portions. 
It  should  be  weighed  between  watch  glasses  in  order  to 
prevent  loss  of  moisture  during  weighing.  Water  is 
estimated,  as  described  under  (a) ,  by  heating  the  sample 
and  determining  the  loss  in  weight.  For  the  estimation 
of  unsaponified  fat  and  other  matter  soluble  in  petroleum 
ether,  such  as  petroleum  or  coal  tar  products,  the  dried 
sample,  as  obtained  from  the  above  operation,  is  extracted 
with  petroleum  ether,  as  described  under  (b).  For  the 
estimation  of  the  total  fatty  or  rosin  acids,  the  material 
from  the  previous  operation  is  exhausted  with  hot  water, 
and  the  liquid  filtered  ;  the  filtrate  containing  in  solution 
the  soap  and  free  alkali  is  treated  with  a  known  volume 
of  normal  nitric  acid,  slightly  in  excess  of  that  required 
to  precipitate  the  whole  of  the  fatty  or  rosin  acids  ;  the 
latter  are  separated  and  estimated,  as  described  under  (c), 
and  further  examined,  as  described  under  (/)  and  (m). 

If,  as  is  often  the  case,  the  amount  of  petroleum  ether 
soluble  material  is  small,  the  extraction  with  petroleum 
ether  may  be  omitted,  and  operation  (c)  may  be  com- 
menced, either  with  the  dried  material  from  the  water 
determination,  or  with  a  fresh  quantity  of  the  soap. 


Soap  1 29 

The  aqueous  solution  is  filtered,  if  necessary,  prior  to  the 
acidification  with  nitric  acid. 

The  liquid  from  which  the  fatty  or  rosin  acids  have 
been  precipitated  and  removed  is  now  titrated  with 
standard  potassium  or  sodium  hydroxide  solution,  using 
methyl  orange  as  indicator  ;  from  the  number  of  cubic 
centimetres  of  alkali  solution  required  to  produce 
neutrality,  and  the  amount  of  normal  nitric  acid  pre- 
viously added,  the  total  alkali  of  the  soap  may  be  calcu- 
lated ;  this  operation  is  described  under  (d).  The 
neutralised  liquid  thus  obtained  is  made  up  to  a  definite 
volume  and  divided  into  five  aliquot  portions  in  which 
the  following  constituents  may  be  determined  as  de- 
scribed under  the  respective  headings  :  soluble  fatty 
acids  (from  cocoanut  or  palm  kernel  oils)  (e),  chlorides 
and  sulphates  (/),  glycerol  (g),  sugar  and  glycerol  (h). 

For  the  estimation  of  the  remaining  constituents  a 
fresh  sample  is  taken  and  extracted  with  absolute 
alcohol ;  the  filtered  alcohol  will  contain  in  solution  the 
soap  proper,  and  any  free  caustic  alkali  which  may  have 
been  present,  while  the  alkali  carbonate  or  silicate,  sand, 
sawdust,  chalk,  etc.,  will  remain  on  the  filter.  This 
operation,  including  the  estimation  of  the  free  caustic 
alkali  in  the  alcoholic  solution  by  titration,  is  described 
under  (i).  The  residue  on  the  filter  is  exhausted  with 
water,  and  the  alkali  carbonate,  silicate,  borate  or 
aluminate,  determined  in  the  solution  as  described 
under  (/).  The  insoluble  residue  is  examined  as  de- 
scribed under  (k). 

The  various  operations  which  have  been  outlined 
above  are  described  in  detail  below.  Subsequently,  the 
estimation  of  phenols  and  neutral  hydrocarbons  in  soap 
will  be  described,  and  finally  the  analysis  of  phenolic 
disinfectants  containing  soap,  such  as  lysol  and  creoline, 
will  be  dealt  with. 

I.O.A.  Q 


130      Industrial  Organic  Analysis 

(a)  Water. — The  following  method  is  due  to  Watson 
Smith  :  5  to  10  grams  of  the  soap,  reduced  to  fine 
shavings,  are  placed  in  a  large  porcelain  crucible  which  is 
set  in  a  sand  bath,  heated  by  a  small  Bunsen  flame. 
The  soap  is  continually  stirred  with  a  glass  rod  (weighed 
with  the  crucible)  having  a  rough  jagged  end  to  facilitate 
the  breaking  up  of  the  mass.  The  process  is  usually 
complete  in  20  to  30  minutes  ;  when,  on  removing  the 
flame,  a  piece  of  plate  glass  placed  over  the  dish  no  longer 
collects  moisture,  the  heating  may  be  discontinued,  and 
the  dish  and  contents,  together  with  the  glass  rod, 
allowed  to  cool  and  weighed.  The  loss  in  weight 
represents  water  with  possible,  traces  of  alcohol  and 
essential  oils.  Burning  of  the  soap  must,  of  course,  be 
avoided  ;  this  will,  however,  if  it  occurs,  immediately 
be  noticed  by  the  characteristic  odour  produced.  The 
results  of  this  process  are  stated  to  be  accurate  to  within 
0-25  per  cent.,  which  is  sufficient  for  technical  purposes. 

As  will  be  seen  from  the  table  of  typical  analyses  of  soap 
on  p.  147,  the  percentage  of  water  varies  very  consider- 
ably in  the  different  grades  of  soap.  The  best  toilet 
soaps  may  contain  as  little  as  10  to  13  per  cent,  of  water, 
while  some  of  the  inferior  hydrated  varieties  will  contain 
as  much  as  70  to  80  per  cent.  Soft  soap  usually  contains 
about  35  to  45  per  cent,  of  water,  and  a  good  yellow  soap 
some  15  to  25  per  cent. 

(b)  Unsaponified  Fat  and  other  Matter  Soluble  in 
Petroleum  Ether. — A  weighed  quantity  of  the  soap, 
dried  as  described  above,  is  placed  in  an  extraction 
thimble  of  fat-free  filter  paper,  and  exhausted  with 
petroleum  ether  which  is  completely  volatile  below  80°, 
in  a  Soxhlet  apparatus.  (See  p.  66.)  The  petroleum 
ether  extract  is  evaporated  to  dryness  in  a  tared  flask, 
'dried  at  100°  and  weighed.  In  addition  to  extraneous 
unsaponifiable  matter,  such  as  vaseline,  coal  tar  pro- 


Soap  131 

ducts,  etc.,  the  residue  thus  obtained  will  contain  any 
unsaponified  fat  or  free  fatty  acids  which  may  be  present, 
together  with  traces  of  essential  oils  added  as  perfumes, 
and  unsaponifiable  matter  occurring  naturally  in  the 
fats.  Unless,  however,  notable  amounts  of  foreign 
material  have  been  added  to  the  soap,  this  residue  should, 
in  most  cases,  be  very  small.  The  estimation  of  phenols 
and  neutral  oils  in  soap  is  described  in  a  subsequent 
portion  of  this  chapter.  In  Allen's  "  Commercial 
Organic  Analysis,"  1911  edition,  Vol.  II.,  p.  425,  will  be 
found  a  systematic  scheme  for  the  examination  of  the 
petroleum  ether  soluble  material  from  soap. 

(c)  Fatty  and  Rosin  Acids. — As  pointed  out  above,  if 
the  soap  does  not  contain  any  appreciable  amount  of 
petroleum  ether  soluble  material,  and  the  last  opera- 
tion has  been  omitted,  the  operation  now  to  be 
described  may  be  commenced,  either  with  a  fresh 
portion  of  the  original  sample  or  the  residue  from 
the  water"  determination.  The  material,  if  previously 
treated  with  petroleum  ether,  is  spread  out  and  warmed 
gently  so  that  the  solvent  may  evaporate  ;  it  is  exhausted 
with  boiling  water,  and  the  aqueous  solution  filtered  or 
decanted  from  any  insoluble  matter  that  may  be  present. 
Normal  nitric  acid  is  then  added  from  a  burette  until  no 
further  precipitate  is  formed  ;  about  10  to  20  c.c.  of  the 
acid  are  further  added,  and  the  total  quantity  used  is 
noted.  The  precipitated  acids  are  allowed  to  solidify, 
and  the  aqueous  liquid  is  decanted  off  and  preserved  for 
further  examination.  The  acids  are  re- melted  and  mixed 
with  hot  water,  allowed  to  solidify  and  the  water  decanted 
off  and  added  to  the  main  portion ;  this  process  is 
repeated  two  or  three  times,  after  which  the  acids  are 
filtered  off  and  washed  with  cold  water  until  the  washings 
are  no  longer  acid  to  methyl  orange.  The  washings  are 

9—2 


132      Industrial  Organic  Analysis 

added  to  the  main  portion  of  the  aqueous  liquid,  which 
will  contain  the  total  alkali  of  the  soap  as  nitrate; 
generally  also  small  quantities  of  chloride  and  sulphate, 
soluble  fatty  acids  if  the  soap  is  derived  from  cocoanut 
or  palm  kernel  oils ;  glycerol  in  the  case  of  soft  soap  or 
soap  made  by  the  cold  process ;  and  any  sugar,  glycerol, 
dextrin,  gelatine  or  other  soluble  foreign  matter  which 
may  have  been  added  to  the  soap. 

The  funnel  with  the  filter  containing  the  insoluble 
acids  is  placed  over  a  small  weighed  beaker  in  an  air 
oven,  and  the  whole  is  heated  to  about  100°.  As  the 
filter  dries,  the  bulk  of  the  melted  acids  will  pass  through 
it  into  the  beaker  below,  where  they  may  be  weighed 
after  cooling.  The  acids  adhering  to  the  filter  and  funnel 
a  e  washed  by  means  of  petroleum  ether,  which  is  com- 
pletely volatile  below  80°,  into  a  tared  flask,  dried  at 
100°  after  removal  of  the  solvent  by  evaporation,  and 
weighed,  their  weight  being  added  to  that  of  the  main 
portion.  The  weight  of  the  total  fatty  acids,  multiplied 
by  0-97,  gives  the  weight  of  the  acid  anhydrides  (or 
radicles)  existing  in  the  soap. 

If  a  weighed  portion  of  the  acids  is  dissolved  in  neutral 
alcohol  and  titrated  with  standard  potassium  or  sodium 
hydroxide  solution,  using  phenol  phthalein  as  indicator, 
the  total  combined  alkali  existing  as  soap  proper  may  be 
found  ;  this  is  calculated  to  Na^O  or  K2O.  The  latter 
may  also  be  found  by  subtracting  the  free  alkali  from 
the  total  alkali,  determined  as  described  below.  The 
soluble  fatty  acids  in  the  aqueous  liquid,  if  any,  are 
determined  as  described  under  (e)  and  added  to  the  main 
portion  in  the  final  statement  of  the  results  of  the  analysis. 

The  valuation  of  soaps  is  largely  based  on  the  per- 
centage of  fatty  acids,  calculated  to  anhydrides,  as 
indicated  above.  The  best  toilet  soaps  will  contain 


Soap  133 

some  80  per  cent,  of  fatty  acids,  good  household  soaps, 
60  to  65  per  cent,  of  fatty  or  fatty  and  rosin  acids,  while 
the  cheaper  grades  which  contain  filling  material, 
sodium  carbonate  or  silicate,  sand,  etc.,  and  the  soft  and 
hydrated  soaps,  may  contain  as  little  as  10  to  20  per  cent, 
of  fatty  or  fatty  and  rosin  acids. 

(d)  Total  Alkali. — The  total  aqueous  liquid  separated 
from  the  fatty  and  rosin  acids,  including  the  filtrate  and 
washings  obtained  as  described  under  (c),  is  titrated  with 
semi-normal  potassium  or  sodium  hydroxide  solution, 
using  methyl  orange  as  indicator.  The  difference  between 
the  number  of  cubic  centimetres  of  normal  nitric  acid  used 
for  the  liberation  of  the  fatty  acids  as  described  under  (c), 
and  half  the  number  of  cubic  centimetres  of  semi-normal 
alkali  used  in  the  last  titration,  will  be  the  number  of 
cubic  centimetres  of  normal  acid  equivalent  to  the  total 
alkali  of  the  soap.     This  is  calculated  to  Na^O  or  K2O,  as 
the  case  may  be. 

(e)  Soluble  Fatty  Acids. — This  estimation  is  based  on 
the  fact   that  the  fatty  acids  show  an   acid  reaction 
towards    phenol    phthalein    but    not    towards    methyl 
orange  ;    the  latter  indicator  may  therefore  be  used  for 
the   estimation   of   free   mineral   acid   by   titration,   in 
presence    of    fatty    acids    which    may    then    be    esti- 
mated by  adding  phenol  phthalein  to  the  solution  which 
has  previously  been  rendered  neutral  towards  methyl 
orange,  and  continuing  the  titration  with  caustic  alkali 
solution   until   a  permanent   pink   tint   appears.     The 
determination  of  the  soluble  fatty  acids  is  accordingly 
carried  out  by  adding  phenol  phthalein  to  the  solution 
from  the  determination  of  the  total  alkali  (as  described 
under  (d)),  and  titrating  with  deci-normal  caustic  alkali 
solution.     The  number  of  cubic  centimetres  required  to 
produce  a  permanent  pink  tint  is  calculated  to  caprylic 


134      Industrial  Organic  Analysis 

anhydride,  rVr15  ^O,   2NaOH  being  equivalent   to  one 


molecule  of  the  anhydride  ;  the  weight  of  the  latter  is 
added  to  the  weight  of  the  anhydrides  of  the  insoluble 
acids  already  estimated  as  described  under  (c). 

The  presence  of  appreciable  amounts  of  soluble  fatty 
acids  will  indicate  that  the  soap  has  been  made  from 
cocoanut  or  palm  kernel  oils. 

(/)  Chlorides  and  Sulphates.  —  These  constituents  may 
be  determined  in  aliquot  portions,  say,  one-fifth  each,  of 
the  neutralised  liquid  obtained  from  operations  (d)  or  (e). 
The  chlorides  are  determined  in  the  usual  manner  by 
titrating  the  faintly  acid  solution  with  deci-normal  silver 
nitrate  solution,  using  a  few  drops  of  potassium  chromate 
solution  as  indicator,  and  continuing  the  titration  until 
a  permanent  faint  red-brown  colour  appears.  The 
number  of  cubic  centimetres  of  silver  nitrate  solution 
required  is  calculated  to  express  the  percentage  of 
NaCl  (or  KC1)  in  the  sample.  Sulphates  are  determined 
by  precipitation  with  barium  chloride  solution,  in  the 
usual  manner,  and  the  amount  of  barium  sulphate 
weighed  is  calculated  to  express  the  percentage  of 
Na^SO4  (or  K2SO4)  in  the  sample. 

As  will  be  seen  from  the  table  on  p.  147,  the  better  the 
quality  of  the  soap,  the  smaller  will  be  the  percentage 
of  sodium  chloride  and  sulphate. 

(g)  Glycerol  (in  absence  of  sugar).  —  The  determination 
of  this  constituent  is  mainly  of  interest  in  the  case  of 
soft  soaps,  and  the  hydrated  soaps  made  by  the  cold 
process.  As  was  pointed  out  above,  some  toilet  soaps 
may  contain  added  glycerol. 

For  the  estimation,  an  aliquot  portion  of  the  neutra- 
lised solution  obtained  from  operations  (d)  or  (e)  may  be 
used.  The  method  to  be  described  is  due  to  Hehner, 
and  depends  on  the  quantitative  oxidation  of  glycerol 
to  carbon  dioxide  and  water  by  means  of  potassium 


Soap  135 

dichromate  solution,  the  amount  of  dichromate  used  for 
the  oxidation  being  measured  by  a  titration  process. 
The  method  is,  of  course,  only  applicable  in  the  absence 
of  other  oxidisable  matter  such  as  sugar,  which,  if 
present,  must  be  removed  as  described  below  under  (ti). 
The  following  solutions  will  be  required  :— 

Potassium  dichromate  solution,  containing  74-56  grams 
of  the  pure  salt  and  about  150  c.c.  of  concentrated 
sulphuric  acid  per  litre. 

Ferrous  ammonium  sulphate  solution,  containing 
240  grams  of  FeSO4(NH4)2SOt6H2O  and  50  c.c.  of  con- 
centrated sulphuric  acid  per  litre. 

Potassium  dichromate  solution  of  exactly  one-tenth  the 
strength  of  the  other  solution  of  the  same  salt.  The  stronger 
dichromate  solution,  i  c.c.  of  which  should  be  equivalent 
to  o-oi  gram  of  glycerol,  is  titrated  against  the  ferrous 
ammonium  sulphate  solution. 

The  solution  in  which  the  glycerol  is  to  be  estimated  is 
made  up  to  about  250  c.c.  and  transferred  to  a  beaker 
which  has  been  cleaned  with  potassium  dichromate  and 
sulphuric  acid,  in  order  to  remove  all  traces  of  oxidisable 
matter.  50  c.c.  of  the  stronger  dichromate  solution  are 
then  added,  and  the  beaker  is  covered  with  a  clock  glass 
and  heated  for  2  hours  in  boiling  water.  At  the  end  of 
this  time  an  excess  of  the  ferrous  ammonium  sulphate 
solution  is  added,  so  that  the  whole  of  the  remaining 
potassium  dichromate  will  be  reduced  to  chromic 
sulphate,  and  a  convenient  amount  of  the  ferrous  salt 
left  over  for  titration  with  the  weaker  dichromate 
solution.  The  approximate  quantities  of  the  various 
solutions  required  may  be  calculated  from  the  following 
equation,  representing  the  interaction  between  the  di- 
chromate and  the  ferrous  salt  in  solution. 
K2Cr207  +  6  FeSO,  +  7  H2SO4  =  K2SO4  +  Cr2(SO4)8 
+  3  Fe2(S04)3  +  7H20. 


136      Industrial  Organic  Analysis 

The  titration  with  potassium  dichromate  solution  is 
carried  out,  using  a  weak  solution  of  potassium  ferri- 
cyanide  as  an  outside  indicator,  according  to  the  well- 
known  method  described  in  text  books  of  quantitative 
inorganic  analysis.  As  previously  stated,  the  glycerol  is 
quantitatively  oxidised  to  carbon  dioxide  and  water  by 
the  potassium  dichromate,  which  is  reduced  to  chromic 
sulphate  as  indicated  by  the  equation  given  above. 
From  these  data  the  amount  of  glycerol  present  in  the 
original  solution  may  be  calculated. 

(h)  Glycerol  and  Sugar  in  presence  of  one  another. — 
Cane  sugar  is  sometimes  added  to  the  cheaper  or  inferior 
grades  of  toilet  soaps,  in  which  it  may  be  present  to  the 
extent  of  20  to  30  per  cent. 

The  estimations  described  under  this  heading  are 
carried  out  on  an  aliquot  portion  of  the  neutralised 
solution  obtained  from  operations  (d)  or  (e).  Carbo- 
hydrates may  be  tested  for  in  a  small  portion  of  this 
solution  by  adding  a  few  drops  of  an  alcoholic  solution  of 
a  naphthol  and  then  i  or  2  c.c.  of  concentrated  sulphuric 
acid  ;  in  the  presence  of  carbohydrates,  a  violet  colora- 
tion will  be  formed  at  the  point  of  contact  of  the  acid  and 
aqueous  layers.  The  presence  of  cane  sugar  may  be 
inferred  if  a  small  portion  of  the  solution  shows  a  marked 
increase  in  its  action  on  Fehling's  solution  (see  "Estima- 
tion of  Cane  Sugar"  below),  after  heating  for  about 
10  minutes  in  boiling  water  with  about  one-tenth  its 
volume  of  concentrated  hydrochloric  acid.  If,  however, 
the  sugar  present  is  glucose  or  invert  sugar,  as  may 
occasionally  be  the  case,  a  pronounced  action  on  Fehling's 
solution  will  be  obtained  at  once.  It  should  be  remem- 
bered that  glycerol  reduces  Fehling's  solution  to  some 
extent  on  prolonged  heating  ;  in  the  presence  of  sugars, 
however,  a  marked  reducing  action  should  be  apparent 


Soap  137 

after  warming  from  half  a  minute  to  a  minute.  Sugar, 
although  oxidisable  by  potassium  dichromate,  cannot  be 
quantitatively  estimated  by  means  of  this  reagent  ;  if, 
therefore,  the  glycerol  is  to  be  estimated  by  the  method 
just  described,  it  must  first  be  separated  from  the 
sugar.  According  to  Donath  and  Meyer,  this  may  be 
accomplished  as  follows  : — 

Add  to  the  solution  slaked  lime  in  sufficient  quantity 
to  combine  with  the  whole  of  the  sugar  present,  allowing 
at  least  3  molecules  of  calcium  hydroxide  to  each  mole- 
cule of  sugar  ;  for  this  purpose,  I  to  2  grams  of  the  pure 
hydroxide  should  be  sufficient  ;  further  add  an  equal 
quantity  of  washed  and  ignited  sand,  boil  the  liquid 
down  to  a  syrup,  pulverise  the  solidified  residue,  and 
exhaust  it  in  a  corked  flask  with  80  to  100  c.c.  of  a 
mixture  of  equal  parts  of  ether  and  alcohol.  The  glycerol 
will  pass  into  solution  while  the  sugar  will  remain 
undissolved.  After  filtering  and  cautiously  removing 
the  solvent  by  evaporation,  the  residual  glycerol  is 
dissolved  in  about  200  c.c.  of  water  and  estimated  as 
described  above  in  the  previous  section  (g).  In  order  to 
avoid  loss  of  glycerol  by  evaporation,  it  should  not  be 
heated  too  strongly  during  the  removal  of  the  solvent. 
On  the  other  hand,  the  alcohol  and  ether,  being  oxidisable 
by  potassium  dichromate,  must  be  entirely  removed. 

Estimation  of  Cane  Sugar. — The  method  commonly 
employed  for  the  gravimetric  estimation  of  sugars  is 
based  on  the  fact  that  the  members  of  this  class,  which 
contain  free  carbonyl  groups  in  the  molecule,  have  the 
power  of  reducing  an  alkaline  solution  of  copper  (see  the 
preparation  of  Fehling's  solution  below)  with  the  pre- 
cipitation of  cuprous  oxide.  The  ratio  between  the 
amount  of  cuprous  oxide  formed  and  the  weight  of  the 
reducing  sugar  cannot  be  expressed  by  definite  chemical 


138      Industrial  Organic  Analysis 

equations,  and,  moreover,  varies  with  the  experimental 
conditions  as  well  as  with  the  amount  of  sugar  taken. 
If,  however,  the  determination  is  carried  out  under 
certain  standard  conditions,  the  amount  of  sugar 
corresponding  to  a  given  weight  of  precipitated  cuprous 
oxide  may  be  ascertained  from  the  tables  of  ratios 
which  have  been  determined  for  each  individual  sugar. 
Cane  sugar  is  a  disaccharide  which  contains  no -free 
carbonyl  group  in  the  molecule  ;  it  consequently  differs 
from  lactose,  maltose,  glucose,  fructose  and  galactose  in 
forming  no  osazone  and  having  no  action  on  an  alkaline 
copper  solution  such  as  Fehling's  solution.  On  heating 
in  aqueous  solution  with  a  little  mineral  acid,  cane  sugar 
is  hydrolysed  with  the  formation  of  glucose  and  fructose 
in  accordance  with  the  following  equation  : — 

CuHujOii  -f-  H2O  =  C6H12O6  -f-  C6H12O6 

Glucose        Fructose 

The  resulting  mixture  is  known  as  invert  sugar,  and  the 
process  is  referred  to  as  the  inversion  of  cane  sugar, 
owing  to  the  change  in  sign  of  the  optical  rotation  of 
the  solution,  from  positive  to  negative,  which  it  occasions. 
A  method  for  the  estimation  of  cane  sugar,  based  on  this 
latter  change,  is  sometimes  employed,  Invert  sugar, 
as  will  be  gathered  from  what  has  already  been  said, 
reduces  Fehling's  solution,  and  the  amount  of  cuprous 
oxide  formed  under  given  conditions  may  be  taken  as  a 
measure  of  the  amount  of  cane  sugar  originally  present 
in  the  solution. 

The  determination  of  cane  sugar  in  soap  may  be 
carried  out  on  an  aliquot  portion  of  the  neutralised 
solution  from  operations  (d)  or  (e).  The  glycerol  need 
not  be  removed,  provided  that  the  solution  is  sufficiently 
dilute,  and  the  heating  with  the  Fehling's  solution  is 
not  unduly  prolonged.  Under  these  conditions  the 


Soap  139 

reducing  action  of  the  glycerol  will  be  negligible,  for  the 
purposes  of  a  technical  analysis. 

The  Fehling's  solution,  which  will  be  required  for  the 
determination,  according  to  the  directions  of  Brown, 
Morris  and  Millar  (/.C.S.,  1897,  Ixxi.,  p.  278),  is  pre- 
pared as  follows  :  173  grams  of  Rochelle  salt  (sodium 
potassium  tartrate)  and  65  grams  of  pure  anhydrous 
.sodium  hydroxide  are  dissolved  in  water  and  made  up 
to  500  c.c.;  34-6  grams  of  pure  crystallised  copper  sul- 
phate, CuSO4  5H2O,  are  dissolved  in  a  separate  quantity 
of  water  and  the  solution  made  up  to  500  c.c.  The  two 
solutions  are  kept  separately,  and  mixed  in  equal 
quantities  when  required  for  use. 

The  inversion  is  carried  out  by  heating  the  sugar 
solution  with  one-tenth  its  volume  of  fuming  hydro- 
chloric acid  for  15  minutes  in  a  water  bath  at  68°  C.,  as 
indicated  by  a  thermometer  placed  in  the  liquid  itself. 

The  amount  of  sugar  used  in  the  determination 
should  be  regulated  so  that  the  amount  of  copper 
weighed  is  from  0-15  to  0-35  gram  (or  the  corre- 
sponding amount  of  cupric  oxide).  For  this,  about 
10  to  30  c.c.  of  a  i  per  cent,  solution  of  invert  sugar  will 
be  required.  In  order  to  ascertain  the  approximate 
amount  of  the  sugar  solution  required,  a  preliminary 
rough  determination  of  its  strength  may  be  made  as 
follows  :  10  c.c.  of  the  mixed  Fehling's  solution  and 
10  c.c.  of  distilled  water  are  placed  in  a  small  flask  and 
heated  to  boiling,  while  the  inverted  sugar  solution  is 
run  in  from  a  burette  until  the  copper  is  completely 
precipitated.  The  end  point  of  the  titration  is  reached 
when  the  blue  colour  has  just  been  discharged,  and  a  drop 
of  the  clear  solution  placed  on  a  piece  of  pure  filter 
paper  beside  a  drop  of  potassium  ferrocyanide  solution 
acidulated  with  acetic  acid  produces  no  brown  stain 


140      Industrial  Organic  Analysis 

at  the  point  where  the  liquids  meet.  The  titration 
should  be  performed  as  rapidly  as  possible  to  avoid  the 
cuprous  oxide  redissolving  owing  to  atmospheric  oxida- 
tion ;  it  is,  therefore,  best  to  verify  the  first  titration  by 
a  second,  in  which  practically  the  whole  of  the  sugar 
solution  required  is  added  at  once,  and  the  solution 
boiled  for  2  minutes  before  testing  for  copper.  Having 
thus  roughly  determined  the  strength  of  the  sugar 
solution  relatively  to  the  Fehling's  solution,  the  latter  is 
standardised  in  exactly  the  same  way  as  just  described, 
against  a  ^  per  cent,  solution  of  cane  sugar  which  has 
been  inverted  as  described  above. 

The  actual  determination  is  carried  out  as  follows  : 
25  c.c.  of  each  of  the  Fehling's  solutions  are  placed  in  a 
beaker  of  about  200  c.c.  capacity,  and  the  mixture  is 
diluted  with  distilled  water  so  that  after  the  addition 
of  the  sugar  solution  the  total  volume  will  be  100  c.c. 
The  beaker  is  covered  with  a  clock  glass  and  heated  in 
boiling  water.  After  a  few  minutes  an  accurately 
measured  portion  of  the  sugar  solution  is  added,  the 
heating  is  continued  for  exactly  12  minutes,  and  the 
precipitated  cuprous  oxide  filtered  off  on  a  Gooch 
crucible  without  delay.  The  precipitate  is  well  washed 
with  hot  water,  then  with  a  little  alcohol,  followed  by 
ether,  and  dried  in  an  oven.  The  cuprous  oxide  is  then 
ignited  to  cupric  oxide  in  the  Gooch  crucible,  the  latter 
being  placed  in  an  ordinary  crucible  which  is  heated 
over  a  Bunsen  flame.  The  cupric  oxide  is  weighed,  and 
converted  to  sugar  by  referring  to  the  table  on  p.  141. 
A  blank  test  should  be  made  on  50  c.c.  of  the  mixed 
Fehling's  solution,  and  any  cuprous  oxide  formed  here 
deducted  from  the  amount  weighed  in  the  actual 
determination. 

From    the    accompanying    table,    the    amounts    of 


Soap 


141 


glucose,  fructose  or  invert  sugar  corresponding  to  given 
weights  of  copper  may  be  found.  Intermediate  values 
may  be  arrived  at  by  interpolation.  The  weight  of 
invert  sugar,  multiplied  by  0-95,  will  give  the  correspond- 
ing weight  of  cane  sugar.  The  factor  for  converting 
CuO  to  Cu  is  07989. 

In  some  cases  it  may  happen  that  the  soap  contains 
nvert  sugar  or  glucose.  These  may  be  determined 
exactly  as  described  above,  omitting  the  heating  with 
hydrochloric  acid.  Any  increase  in  the  amount  of 
cuprous  oxide  precipitated  after  treatment  with  hydro- 
chloric acid  would  be  due  to  cane  sugar. 

TABLE  FOR  FINDING  AMOUNTS  OF  GLUCOSE,  AND  INVERT 
SUGAR,  CORRESPONDING  TO  GIVEN  WEIGHTS  OF  COPPER, 
OBTAINED  BY  THE  METHOD  OF  BROWN,  MORRIS  AND  MlLLAR 
(Journ.  Chem.  Soc.,  Vol.  71,  p.  281). 


Glucose 
Grams. 

Copper 
Grams. 

Glucose 
Grams. 

Copper 
Grams. 

Invert 
Sugar 
Grams. 

Copper 
Grams. 

Invert 
Sugar 
Grams. 

Copper 
Grams. 

0-050 

0*1030 

0-155 

0-3020 

0-050 

0-0975 

0-155 

0-29I5 

0-055 

0-1134 

0-160 

0-3103 

0-055 

0-1076 

o-i6o 

0-3002 

0-060 

0-1238 

0-165 

0-3187 

0-060 

0-1176 

0*165 

0-3086 

0-065 

0-1342 

0-170 

0-3268 

0-065 

0-1275 

0-170 

0-3167 

0/070 

0-1443 

0-175 

0-3350 

0-070 

°'I373 

0-175 

0-3251 

0-075 

0-1543 

o'i8o 

0-343I 

0-075 

0-1468 

0-180 

03331 

0-080 

0*1644 

0-185 

0-3508 

0-080 

0-1566 

0-185 

0-3410 

0-085 

0-1740 

0-190 

0-3590 

0-085 

0-1662 

0-190 

0-3490 

0-090 

0-1834 

0-195 

0-3668 

0-090 

0-1755 

o-J95 

0-357° 

0-095 

0-1930 

O'2OO 

0-3745 

0-095 

0*1848 

0-200 

0-3650 

O'lOO 

0-2027 

0-205 

0-3822 

OTOO 

0-1941 

0-205 

0-3726 

0-105 

0*2123 



—  • 

0-105 

0-2034 

O'lIO 

0'22l8 



— 

O'HO 

O-2I28 

0-115 

0-23I3 



— 

0-115 

O'2220 

0-120 

0-2404 



— 

O'I20 

0-23II 

O-I25 

0-2496 



— 

OT25 

0-2400 

O-I30 

0-2585 

• 

— 

OT3O 

0-2489 

°'I35 

0-2675 



— 

0-135 

0-2578 

0-140 

0*2762 



— 

0-140 

O-2662 

P'i45 

0-2850 



— 

OT45 

0-2750 

0-150 

0-2934 

~~ 

OT50 

0-2832 

142      Industrial  Organic  Analysis 

Although  results  of  very  great  accuracy  are  not 
required  in  the  present  instance,  the  method  has  been 
described  in  detail  as  it  will  be  referred  to  in  connection 
with  work  to  be  described  in  subsequent  chapters. 

(i)  Free  Caustic  Alkali  or  Fatty  Acids. — According  to 
Hope,  10  to  30  grams  of  the  original  sample  are  dissolved 
in  hot  absolute  alcohol  in  a  flask,  loosely  corked  to 
prevent  absorption  of  moisture  and  carbon  dioxide  from 
the  air.  If  the  soap  contains  much  water,  it  should 
first  be  partially  dried  in  an  atmosphere  free  from  carbon 
dioxide.  The  hot  solution  is  filtered  rapidly  to  prevent 
undue  exposure  of  the  solution  to  air,  care  being  taken 
that  none  of  the  soap  jelly  separates  out  on  the  filter. 
The  filter  is  washed  with  absolute  alcohol,  and  the  total 
filtrate  and  washings  titrated  in  presence  of  phenol 
phthalein  with  deci-normal  acid  or  alkali,  according  to 
its  reaction.  If  alkaline,  the  amount  of  standard  acid 
required  will  give  the  percentage  of  free  caustic  alkali 
in  the  soap  ;  if  acid,  the  amount  of  alkali  required  is 
calculated  to  oleic  acid  (C18H34O2)  and  returned  as  free 
fatty  acids. 

Toilet  soaps  should  be  entirely  free  from  caustic 
alkali,  while  in  laundry  soaps  the  percentage  of  this 
constituent  should  not  be  above  J  per  cent. 

By  titrating  the  neutralised  alcoholic  solution  with 
acid,  using  this  time  methyl  orange  as  indicator,  the 
alkali  existing  in  the  soap  as  sodium  salts  of  fatty  or 
rosin  acids  may  be  found.  (See  also  (d).) 

(/)  Alkaline  Carbonate,  Silicate,  etc. — The  residue  left 
on  the  filter  contains  any  carbonate,  silicate,  borate, 
aluminate,  etc.,  together  with  other  insoluble  material 
such  as  starch,  sawdust,  sand,  pumice,  chalk,  insoluble 
colouring  matter,  etc.,  which  may  have  been  present  in 
the  soap.  The  alkaline  carbonate,  silicate,  borate  or 


Soap  143 

aluminate  is  separated  from  the  residue  insoluble  in 
alcohol  by  extracting  with  cold  water  and  filtering.  One 
half  of  the  filtrate  may  then  be  titrated  with  deci-normal 
sulphuric  acid  in  presence  of  methyl  orange  as  indicator, 
in  order  to  find  the  amount  of  soda  present,  while  the 
other  half  may  be  evaporated  to  dryness  and  the  nature 
of  the  alkaline  material  determined  by  examining  the 
residue  by  the  usual  methods  of  qualitative  inorganic 
analysis. 

In  a  good  toilet  soap,  the  amount  of  alkaline  carbonate 
should  not  exceed  J  per  cent.  In  laundry,  scouring 
and  marine  soaps,  the  addition  of  larger  amounts  of 
alkaline  carbonate  or  silicate  may  be  quite  permissible. 

(k)  Insoluble  Residue  from  (/). — The  residue  left  on 
the  filter  after  extracting  the  alkaline  material  with 
water,  may  be  dried,  weighed  and  further  examined. 
The  substances  mentioned  above  (p.  127)  will  readily  be 
recognised  by  the  usual  tests,  as  well  as  their  appearance. 

Total  Alkali  and  Nature  of  the  Same. — By  titrating  an 
aqueous  solution  of  a  known  weight  of  the  soap  with 
standard  acid,  using  methyl  orange  as  indicator,  the 
total  alkali  present  in  the  soap  may  be  found.  This 
should  agree  with  the  sum  of  the  determinations  of 
alkali  as  (i)  free  hydroxide,  (2)  free  carbonate,  or  silicate 
borate,  etc.,  and  (3)  combined  as  soap  with  fatty  or  rosin 
acids. 

If  it  is  desired  to  determine  the  nature  of  the  alkali, 
i.e.,  whether  soda  or  potash  or  a  mixture  of  these,  the 
alcoholic  soap  solution  neutralised  to  methyl  orange 
from  operation  (7)  is  treated  with  baryta  solution  until 
alkaline  to  phenol  phthalein,  and  barium  chloride  solu- 
tion is  added  as  long  as  precipitation  occurs.  After 
filtering,  the  solution  is  evaporated  to  dryness,  and  the 
residue  examined  qualitatively  or  quantitatively  for 


144      Industrial  Organic  Analysis 

sodium  or  potassium  by  any  of  the  usual  methods 
described  in  the  text-books  on  qualitative  and  quantita- 
tive inorganic  analysis. 

(/)  Detection  and  Estimation  of  Rosin  Acids. — Rosin 
acids  may  be  detected  by  the  Liebermann  Storch  reaction 
as  follows  :  a  small  portion  of  the  acids,  precipitated  and 
separated  as  described  under  (c)  is  dissolved  in  acetic 
anhydride  at  a  gentle  heat,  and  the  solution  allowed  to 
cool.  Sulphuric  acid  of  specific  gravity  1-53,  prepared 
by  adding  347  of  the  concentrated  acid  to  37-5  c.c.  of 
water,  is  very  carefully  added  when,  in  the  presence  of 
rosin  acids,  a  reddish  violet  coloration  will  be  developed. 
If  the  acid  is  not  added  with  sufficient  care,,  the  mixture 
will  become  too  warm,  with  the  result  that  the  colora- 
tion will  not  be  observed,  a  brownish-yellow  coloration 
being,  developed  at  once.  Lewkowitsch  points  out  that 
cholesterol  gives  a  similar  reaction  to  that  obtained  with 
rosin  acids,  and  recommends  that  if  the  presence  of  this 
substance  is  suspected,  it  should  be  removed  by  extract- 
ing the  aqueous  soap  solution  with  ether  before  liberating 
the  acids  for  the  test.  As  has  been  pointed  out  in  the 
previous  chapter  on  the  fatty  oils  and  fats,  cholesterol 
is  the  principal  unsaponifiable  constituent  of  the  animal 
fats. 

For  the  estimation,  the  mixture  of  fatty  and  rosin 
acids  isolated  as  described  under  (c),  may  be  used.  The 
method  of  Twitchell  depends  on  the  fact  that  while  the 
fatty  acids  are  esterified  on  treatment  with  hydrogen 
chloride  in  alcoholic  solution,  the  rosin  acids  are  un- 
changed, and  may  be  estimated  by  titration  with 
standard  alkali  solution  in  presence  of  phenol  phthalein 
as  indicator. 

2  to  3  grams  of  the  mixed  fatty  and  rosin  acids  are 
dissolved  in  10  times  their  volume  of  alcohol,  in  a  flask, 


Soap  145 

and  dry  hydrogen  chloride  gas  is  bubbled  through  the 
solution  until  no  further  absorption  takes  place.  The 
process  will  be  complete  in  about  45  minutes,,  when  the 
ethyl  esters  of  the  fatty  acids  will  have  separated  as  an 
oily  layer  floating  on  the  alcohol.  The  liquid  is  now 
diluted  with  five  times  its  volume  of  water  and  boiled 
till  the  acid  solution  is  clear,  the  esters  containing  the 
rosin  acids  in  solution  floating  on  the  top.  The  whole 
is  transferred  to  a  separating  funnel  by  means  of  ether, 
and  the  ethereal  layer  is  washed  with  water  till  'all  the 
mineral  acid  has  been  extracted  from  it ;  this  will  be 
accomplished  when  the  aqueous  layer  no  longer  reacts 
acid  towards  methyl  orange,  or  gives  a  precipitate  or 
opalescence  with  silver  nitrate  solution  in  presence  of 
nitric  acid.  The  ethereal  layer  is  now  transferred  to  a 
flask,  50  c.c.  of  alcohol,  neutralised  towards  phenol 
phthalein,  are  added,  and  the  mixture  is  titrated  with 
standard  sodium  hydroxide  solution,  using  phenol 
phthalein  as  indicator.  The  titration  should  not  be 
unduly  prolonged,  or  partial  hydrolysis  of  the  esters  by 
excess  of  alkali  may  take  place.  From  the  number  of 
cubic  centimetres  of  alkali  solution  required  for  neutra- 
lisation, the  amount  of  the  rosin  acids  may  be  calculated, 
assuming  their  combining  .weight  to  be  346. 

As  mentioned  above,  rosin  should  only  be  present  in 
the  cheaper  grades  of  soap  ;  a  good  laundry  soap  should 

only  contain  about  20  per  cent,  of  rosin  acids. 

• 

(m)  Origin  of  the  Fatty  Acids. — It  will  not  often  be 
necessary  to  determine  the  nature  of  the  fat  from  which 
the  soap  is  derived.  As  mentioned  above,  if  an  appreci- 
able quantity  of  soluble  fatty  acids  is  found  by  the  pro- 
cess described  under  (e) ,  the  soap  probably  contains  fatty 
acids  derived  from  cocoanut  or  palm  kernel  oils.  In 
some  cases  it  may  be  possible  to  determine  the  origin  of 
the  fatty  acids  by  determining  such  constants  as  the 

I.O.A.  10 


146      Industrial  Organic  Analysis 

solidifying  point  (liter  test),  saponification  value,  iodine 
value  and  specific  gravity,  as. described  in  Chapter  III. 
The  problem  is,  however,  usually  not  an  easy  one,  and 
becomes  decidedly  complicated  if  rosin  acids  are  present, 
or  mixtures  of  fats  have  been  used.  Further,  the  fats 
used  in  the  manufacture  of  the  soap  may  in  some  cases 
have  been  bleached  by  potassium  dichromate  and 
hydrochloric  acid  or  similar  mixtures,  when  the  value  of 
the  indications  afforded  by  the  methods  just  alluded  to, 
especially  the  iodine  value,  will  be  considerably  lessened. 
The  Results  of  the  Analysis. — The  accompanying  tables 
contain  typical  analyses  of  various  grades  of  soap,  most 
of  which  are  by  C.  Hope.  As  a  rule,  the  most  important 
of  the  determinations  described  above  are  those  of  water, 
alkali  combined  as  soap,  to  be  calculated  as  Na^O,  free 
caustic  alkali  as  NaOH  (or  KOH),  sodium  (or  potassium) 
carbonate,  water  insoluble  material,  i.e.,  fillers  or 
adulterants  (see  under  (&)),  fatty  acids,  calculated  as 
anhydrides,  and  rosin  acids. 

THE    DETERMINATION    OF    PHENOLS    AND    NEUTRAL 
HYDROCARBONS  IN  CARBOLIC  SOAPS. 

The  following  method  for  the  analysis  of  carbolic 
soaps  is  due  to  Allen  :  5  grams  of  the  sample  are  dis- 
solved in  warm  water,  and  20  to  30  c.c.  of  a  10  per  cent, 
solution  of  sodium  hydroxide,  according  to  the  amount 
of  phenols  supposed  to  be  present,  are  added.  The  alka- 
line solution  is  transferred  to  a  separating  funnel  and 
extracted  with  ether.  The  ethereal  solution  is  evapo- 
rated in  a  tared  flask  on  the  water  bath,  and  the  residue 
weighed  ;  this  residue,  which  will  consist  of  neutral  oils 
of  tar  or  other  substances  of  a  like  nature  which  may  have 
been  added  to  the  soap,  may  be  appreciably  volatile 
at  100°,  in  which  case  it  will  be  difficult  to  get  an  accurate 
estimate  of  its  amount. 

The  alkaline  liquid  which  has  been  separated  from  the 
ether  is  then  transferred  to  a  more  capacious  separating 


Soap 


147 


"d     . 

• 

C  —  ' 

_• 

_j 

•^ 

—• 

C  'S 

'm  'S 

r, 

o 

_j 

c 

,   O 

'35   c 

2c 

<D 

05 

— 

o 

fH 

d 

c 

C 

r^J      ^ 

S 

c^ 

1 

0 

JD 

-M 

-"  O 
>.  O 

^•d 

I'd 

0 

0    ^4 

s  a 

^  d 

11 

'd 

-^ 
S 

1 

o3 

o3   U 

c3   ^ 

CTJ 

H 

03     03 

03   h 

3 

H 

H 

H  rt 

H 

U 

PLH     fin 

H 

PL4 

o 

•<*• 

o 

o 

0 

00       O 

oo 

iO 

o 

0) 

*j  "2  "$,  -o 

04 

o 

iO 

o 

O        M 

l>« 

M 

04 

O 

1 

ctf   <tf   O   O 

H 

iO 

ro 

w 

M- 

M        0 

O4 

M 

M 

0 

1 

UH        &< 

^ 

^ 

10 

o 

O      Tf 

o 

* 

M 

0< 

. 

00 

£ 

oo 

04 

CO 

M 

O     ON 

00        M 

? 

a 

» 

-h 

55 

i 

8 

% 

S 

ft 

8 

M 

ON     ON 

0 

o 

H  1 

85 

ft 

C^ 

8 

M 

o 

2- 

Jf 

5 

00 

O 

0) 

in   o 

ol 

0 

"*• 

O 

o 

01 

10 

04 

ffi 

,1, 

oo 

{^ 

oo 

oo 

04     00 

M 

ri 

-(- 

rn 

ro 

04 

M 

04 

ro     ro 

ro 

10 

00 

w> 

M 

,v 

^ 

rf 

o 

O 

O 

O      ro 

04 

10 

^. 

O 

ON 

rt  C  ^ 

O 

.  ro 

M 

M 

M      O 

O 

M 

M 

H 

O 

0 

O 

O 

o 

o 

O      O 

O 

O 

0 

0 

0 

S<5 

o 

ro 

o 

g 

c 

ro    04 

M         t^. 

04 

CO 
H 

o 

rrj 

^ 

o 

O 

o 

o 

O      O 

o 

O 

0 

0 

_,• 

ON 

04 

0 

H 

0 

!>.  CO 

o 

00 

^ 

CO 

•o 

rt 

Tj- 

ro 

iO 

^     t*>- 

*!• 

-f- 

iO 

M 

55 

O 

O 

o 

o 

o 

O        H 

0 

M 

04 

•0 

o 

iii 

04 

iO 

iO 

ro 
ro 

£ 

ro   O 

o 

M 

o 

M 

o 

1 

g 

£  rt^ 

0 

0 

0 

0 

o 

O        M 

o 

o 

ro 

O       ,   <u 

o 

oo 

M 

ro 

M        ON 

•^> 

10 

"•» 

on 

^ 

ro 

Tf 

0 

0 

0       04 

2 

M 

'8 

cr* 

1 

£     £3 

S 

04 

o 

M 

O 

O        M 

o 

ro 

6 

(M 

o 

04 
O 

cT 

% 

$ 

04       O 
•«*•     ^t- 

Tf 

o 

0 

o5 

8 

I 

53 

o 

^ 

M 

04 

o 

O     O 

o 

o 

o 

ON 

O      o, 

<N    U"'     Cv 

o 

ON 

04' 

co 

ON 

8 

N 

iO   O 

04 
04 

o 

ro 

H 

J 

£       Crt 

CO 

O 

^ 

j> 

'" 

0     »0 

^ 

-f 

M 

ro 

ON 

O 

jj^ 

o 

«o 

O 

00      ON 

O 

IN 

O 

M 

^•C  -jj  >"•  <u 

o 

04 

ro 

ON 

0 

04     00 

ON 

ON 

ON 

t>. 

"5  §  °  "c  -^ 

ON 

^j. 

M 

ON 

M 

ON    00 

Q 

ON 

O 

ON 

O 

fa      «<TJ 

O 

Tf 

^ 

^ 

vp 

•O     CO 

0 

ro 

M 

M 

^ 

• 

' 

t—I 

I—  J 

i—  i 

• 

•        • 

• 

,_; 

(D 

c 
_o 

ft, 

1 

Q 

H-  1 

2 

M 

I 

\-i 

<D 

« 
1 

<u 
oS 

2 
o 

u 

Marseilles  I 

t—  I 

'o   ^ 
fi    ^3 
J5    -^ 

d 

)H 

15 
fk 

Pale  rosin  I 

"  Yellow  " 
foreign  mar 

Marine 

emigrants 

O 
1 

148      Industrial  Organic  Analysis 

funnel  and  treated  with  an  excess  of  saturated  brine, 
which  will  precipitate  the  soaps  while  the  phenols 
remain  in  solution.  After  agitating  to  coagulate  the 
soap,  the  liquid  is  filtered  ;  in  case  the  soap  should  not 
coagulate  easily,  the  addition  of  a  little  tallow  or  palm 
oil  soap  dissolved  in  water,  will  often  facilitate  the 
separation.  The  precipitated  soap  is  washed  twice  with 
strong  brine,  and  the  washings  are  added  to  the  main 
filtrate  which  is  then  diluted  to  I  litre.  100  c.c.  of 
this  solution,  which  will  be  equivalent  to  0-5  gram  of 
soap,  are  placed  in  a  500  c.c.  stoppered  bottle  and 
acidulated  with  sulphuric  acid,  when  it  should  remain 
perfectly  clear  ;  if  precipitation  occurs  at  this  stage,  the 
fatty  acids  will  not  have  been  completely  removed.  In 
such  cases  200  c.c.  of  the  alkaline  liquid  should  be  taken, 
and  powdered  salt  dissolved  in  it  to  saturation  ;  the 
solution  thus  obtained  should  then  be  filtered  through 
a  dry  filter,  and  100  c.c.  of  the  filtrate  acidified  as  before. 
The  phenol  is  then  determined  by  titration  with  standard 
bromine  water  (see  below),  which  is  run  in  from  a 
burette,  the  stopper  of  the  bottle  being  replaced  and 
the  contents  agitated  after  each  addition.  The  end  point 
is  reached  when  the  solution  acquires  a  permanent 
faint  yellow  tint.  If  crystallised  phenol  has  been 
introduced  into  the  soap  the  precipitated  tribromophenol 
will  be  precipitated  in  snow-white  flocks  which  allow  the 
faintest  tint  owing  to  excess  of  bromine  to  be  observed 
quite  readily.  If,  on  the  other  hand,  cresylic  acid,  i.e., 
the  cresols,  has  been  added,  the  precipitate  will  be  milky 
and  will  not  separate  well  from  the  liquid,  though  the 
end  point  can  still  be  observed.  The  addition  to  the 
original  solution  of  a  known  amount  of  pure  phenol  may 
cause  the  precipitate  to  coagulate,  in  which  case  the 
yellow  colour  due  to  excess  of  bromine  will  be  more 


Soap  149 


easily  seen.     This  addition  must,  of  course,  be  allowed 
for  in  calculating  the  result  of  the  analysis. 

The  bromine  solution  which  is  used  for  the  titration 
is  conveniently  prepared  by  mixing  in  a  stoppered 
bottle  one  measure  of  saturated  bromine  water  with 
two  measures  of  water.  The  resulting  solution,  which 
is  approximately  of  i  per  cent,  strength,  should  be 
standardised  immediately  after  use  by  means  of  a 
standard  solution  of  phenol  of  the  quality  indicated  by 
the  analysis  to  be  present  in  the  sample.  The  phenol 
solution  may  be  made  by  dissolving  0-5  gram  of  the 
phenol  (usually  either  Calvert's  No.  2  or  No.  5  carbolic 
acid)  in  20  c.c.  of  a  10  per  cent,  solution  of  sodium 
hydroxide  with  5  grams  of  a  non-phenolic  soap.  The 
solution  is  precipitated  with  brine  in  the  same  way  as 
the  solution  of  the  sample  under  examination,  the  nitrate 
diluted  to  one  litre,  and  100  c.c.  acidulated  with  sul- 
phuric acid  and  titrated  with  the  bromine  water. 

Carbolic  soaps  are  often  sold  as  containing  20  or  30  per 
cent,  of  carbolic  acid  ;  they  may  contain  anything  from 
i  to  30  per  cent,  of  phenol,  derived  either  from  pure 
crystallised  phenol  or  the  common  cresylic  acids. 
According  to  Allen,  the  proportion  of  phenol  found 
is  occasionally  less  than  that  stated  to  be  present,  the 
difference  probably  being  due  to  loss  by  evaporation. 

PHENOLIC  DISINFECTANTS  CONTAINING  SOAP. 

The  preparations  to  be  dealt  with  under  this  heading 
consist  of  creosote  oil  or  coal  tar  phenols,  chiefly  cresols 
(cresylic  acid) ,  to  which  soap  has  been  added  in  order  to 
form  an  emulsion  or  a  homogeneous  product  which  shall 
be  readily  miscible  with  water.  As  examples  we  may 
take  the  preparations  known  as  creoline  and  lysol. 
The  former  consists  of  creosote  oil  and  an  aqueous 
solution  of  soda  rosin  soap,  which  have  been  mixed  to 
form  an  emulsion.  Being  made  from  the  crude  creosote 


150      Industrial  Organic  Analysis 

oil,  it  contains  some  40  to  60  per  cent,  of  neutral  hydro- 
carbons and  about  10  to  20  per  cent,  of  cresylic  acid  ; 
the  better  the  quality,  the  less  of  the  former  and  the 
more  of  the  latter  will  it  contain.  Creoline  should  form 
a  permanent  emulsion  when  mixed  with  water  in  the 
proportion  of  i  to  40,  the  greater  part  of  the  cresylic 
acid  dissolving  as  such,  or  as  sodium  salts,  and  the  neutral 
hydrocarbons  remaining  suspended  in  the  form  of  minute 
globules  owing  to  the  emulsifying  action  of  the  soap. 
Lysol  and  sapocarbol,  on  the  other  hand,  are  prepared 
by  mixing  cresylic  acid  containing  relatively  small 
amounts  of  hydrocarbons,  with  rosin  or  a  fatty  oil  such 
as  linseed  oil,  and  saponifying  the  rosin  or  oil  by  treating 
the  mixture  with  a  solution  of  potash  in  water  and 
alcohol.  The  viscous  brown  transparent  liquid  thus 
obtained  is  readily  soluble  in  water.  Lysol  should  con- 
tain at  least  40  to  50  per  cent,  of  phenols,  and  at  most 
only  about  3  or  4  per  cent,  of  hydrocarbons,  in  order 
that  it  may  dissolve  in  water  as  completely  as  possible. 
6esides  the  products  just  described,  there  are  a  number 
of  other  similar  preparations  on  the  market  containing 
varying  proportions  of  cresylic  acid,  soap  and  hydro- 
carbons. Among  these  may  be  mentioned  sanatol, 
saprol,  solutol  and  salveol.  Many  sheep  dips  are  similar 
in  composition  to  the  above,  and  may  be  analysed  by 
the  methods  given  below.  The  latter  preparations 
sometimes  contain  sodium  arsenate  or  sodium  carbonate 
in  addition  to  or  in  place  of  the  soap. 

The  Analysis  of  Creoline,  Lysol,  Sheep  Dips,  etc. — The 
estimation  of  neutral  oils  of  tar,  coal  tar  phenols,  pyridine 
bases  and  fatty  or  rosin  acids  may  be  carried  out  as 
follows  :  20  grams  of  the  sample  are  made  distinctly 
acid  by  the  addition  of  dilute  sulphuric  acid  (i  to  5),  and 
agitated  with  ether  in  a  separating  funnel  ;  after 
separating,  the  aqueous  layer  is  again  extracted  with 
two  successive  portions  of  ether.  The  pyridine  bases 
may  be  estimated  in  the  acid  liquid,  as  described  in 
Chapter  II.,  under  the  "  Testing  of  Creosoting  Liquor." 


Soap  151 

The  ethereal  extracts,  containing  the  hydrocarbons, 
phenols  and  fatty  or  rosin  acids,  are  united  and  extracted 
with  three  successive  portions  of  50  c.c.  of  a  10  per 
cent,  solution  of  sodium  hydroxide  ;  the  phenols  and 
acids  are  thus  removed  from  the  ether,  which  retains 
only  the  hydrocarbons  in  solution.  The  alkaline 
extracts  are  united  and  extracted  once  or  twice  with 
ether  in  order  to  free  them  from  all  traces  of  suspended 
hydrocarbons  ;  all  the  ethereal  extracts  are  then  united, 
dehydrated  by  means  of  calcium  chloride  and  evaporated 
in  a  small  tared  flask  on  the  water  bath.  The  residue 
is  dried  at  100°  and  weighed  as  hydrocarbons.  The 
alkaline  liquid  is  acidified  with  dilute  sulphuric  acid, 
allowed  to  cool,  and  the  liberated  acids  and  phenols  are 
removed  by  shaking  out  several  times  with  ether.  The 
separation  of  the  phenols  and  the  acids  contained  in 
solution  in  the  ether  is  accomplished  by  shaking  out  the 
latter  with  dilute  sodium  carbonate  solution,  which  will 
extract  only  the  acids  as  sodium  salts.  The  acids  may 
then  be  liberated  by  the  addition  of  dilute  sulphuric  acid 
to  the  alkaline  solution,  and  separated  and  estimated 
as  described  in  the  present  chapter  under  (c).  If  desired, 
they  may  be  examined  for  rosin  acids,  as  described  on 
p.  144.  The  ethereal  extract,  which  has  been  shaken 
out  with  sodium  carbonate  solution  till  all  the  acids 
have  been  removed,  contains  the  phenols  ;  it  may  be 
evaporated  in  a  tared  flask  on  the  water  bath,  and  the 
residue  dried  at  100°  and  weighed. 

Soda  or  potash  may  be  estimated  by  burning  off 
about  5  grams  of  the  original  sample  in  a  platinum 
crucible  or  dish,  treating  the  residue  with  successive 
small  portions  of  concentrated  sulphuric  acid,  and 
weighing  the  ignited  sodium  or  potassium  sulphate  in 
the  usual  manner. 


152      Industrial  Organic  Analysis 

REFERENCES. 

"  The  Chemical  Technology  and  Analysis  of  Oils,  Fats  and 
Waxes,"  by  J.  Lewkowitsch,  3  vols.  1909.  Macmillan  &  Co., 
London. 

Allen's  "  Commercial  Organic  Analysis,"  Vol.  II.  1910. 
J.  and  A.  Churchill,  London. 

"  Engineering  Chemistry,"  by  Stillmann.  1910.  Williams  and 
Norgate,  London,  and  the  Chemical  Publishing  Co.,  Easton,  Pa. 

"  Chemisch-Technische  Untersuchungsmethoden,"  Vol.  III., 
by  Lunge.  1911.  J.  Springer,  Berlin. 

"  Chemisch-Technische  Analyse,"  Vol.  I.,  by  Post.  1908 
Fr.  Vieweg,  Brunswick. 

"  Einheitsmethoden  zur  Untersuchung  von  Fetten,  Oelen, 
Seifen  und  Glyzerinen,  sowie  sonstigen  Materialen  der  Seifenin- 
distrie,"  by  Donath  and  Margosches  (Sammlung  Chemischer  und 
Chemisch-Technische  Vortrage) . 

"  Technologic  der  Fette  und  Oele,"  by  Gustav  Hefter,  3  vols- 
1906  to  1910.  J.  Springer,  Berlin. 

Also  other  works  mentioned  after  Chapter  III. 

"  The  Soap  Maker's  Handbook,"  edited  chiefly  from  the 
German  of  Deite,  Engelhardt  and  Wiltner,  by  W.  Braunt.  1912, 
E.  F.  Spon,  Ltd.,  London  ;  Henry  Carey,  Baird  &  Co.,  Phila- 
delphia. 

"  Traite  de  Savonnerie,"  by  E.  Moride.  1909,  Librairie 
Polytechnique,  Paris. 


CHAPTER   V 

PETROLEUM    AND    ITS     DISTILLATION    PRODUCTS — LUBRI- 
CATING  OILS 

INTRODUCTORY. 

THE  petroleums,  or  mineral  oils,  are  usually  more  or 
less  viscous,  dark  brown  or  black  fluids  having  specific 
gravities  lying  between  0-73  and  0-97.  Chemically,  they 
consist  of  complex  mixtures  of  hydrocarbons  together 
with  smaller  quantities  of  oxygen,  nitrogen  and  sulphur 
compounds ;  the  nature  of  the  hydrocarbons  varies 
with  the  locality  from  which  the  petroleum  is  obtained, 
and,  to  a  lesser  degree,  with  the  age  of  the  well.  Most 
petroleums  distil  over  a  wide  range  of  temperature  and 
yield  a  variety  of  important  products  such  as  motor 
spirit,  burning  oil,  lubricating  oil,  etc.,  while  the  residues 
which  remain  after  distilling  off  the  more  volatile 
portions  have  in  recent  years  found  a  most  important 
use  as  fuel  for  internal  combustion  engines. 

Composition  of  the  Petroleums. — As  examples,  we  may 
take  the  petroleums  of  Pennsylvania,  U.S.A.,  and  the 
Caucasus,  which  together  constitute  by  far  the  greater 
part  of  the  world's  supply  of  this  commodity. 

Pennsylvania  petroleum,  which  usually  has  a  specific 
gravity  lying  between  0-80  and  0-87,  consists  chiefly  of 
paraffins  from  pentane  C5Hi2  (b.p.  38°)  up  to  the  solid 
C2oH42 ;  higher  members  of  the  series,  such  as  C25H52 
and  QsoH^,  are  also  known  to  be  present.  The  portions 
boiling  above  200°  probably  contain  defines  in  addition 
to  the  paraffins  ;  as  is  also  the  case  with  the  other 
petroleums,  many  of  the  higher  boiling  constituents  are 
of  unknown  constitution.  Aromatic  hydrocarbons  are 
present  only  in  small  amount.  Pennsylvania  petroleum 
is  accompanied  by  hydrogen  and  the  members  of  the 


154      Industrial  Organic  Analysis 

paraffin  series  which  are  gaseous  at  the  ordinary  tempera- 
ture. 

Caucasian  petroleum  has  a  higher  specific  gravity  than 
Pennsylvania  petroleum,  usually  from  0-88  to  0-94,  and 
differs  widely  from*the  latter  in  chemical  composition. 
According  to  Markownikoff,  it  contains  at  least  80  per 
cent,  of  naphthenes,  i.e.,  polymethylenes  and  their 
alkyl  derivatives,  among  which  pentamethylene,  hexa- 
methylene,  methyl  hexamethylene  and  other  similar 
compounds  of  known  constitution  have  been  recognised. 
The  naphthenes  resemble  the  paraffins  in  being  saturated 
bodies  which  are  not  acted  on  by  concentrated  sulphuric 
acid  or  potassium  permanganate  ;  with  the  halogens  and 
dilute  nitric  acid,  they  yield  halogen  and  nitro  substitu- 
tion products.  In  addition  to  the  naphthenes,  Russian 
petroleum  contains  about  10  per  cent,  of  aromatic 
hydrocarbons,  and  less  than  i  per  cent,  of  paraffins. 
The  higher  boiling  fractions  appear  to  contain  unsatu- 
rated  substances  such  as  olefines  and  naphthylenes,  the 
latter  being  olefine  analogues  of  the  naphthenes.  Like 
Pennsylvania  oil,  Russian  oil  contains  a  number  of  sub- 
stances of  unknown  constitution,  especially  in  the  higher 
boiling  fractions. 

From  the  foregoing  it  will  be  gathered  that  the  two 
classes  of  petroleums  described  show  important  differ- 
ences in  their  behaviour  on  distillation.  Generally 
speaking,  Pennsylvania  petroleum  distils  over  a  wider 
range  of  temperature,  and  contains  a  much  larger  pro- 
portion of  volatile  matter  than  Russian  petroleum,  the 
residue  from  the  latter  being  larger  in  amount  and  more 
fluid  in  consistency  than  that  from  the  former.  The 
primary  products  of  distillation  of  the  petroleums  are  as 
follows  :  (i)  Light  oils  or  mineral  naphtha,  from  which 
are  obtained  by  fractional  distillation  various  mixtures 
of  light  hydrocarbons  which  are  used  as  fuel  for  motor 
engines,  extraction  and  cleaning  purposes,  and  burning 
naphtha.  (2)  Illuminating  oils,  such  as  kerosene  or 
solar  oil.  (3)  Residuum,  which  in  the  case  of  Russian 
oils  is  generally  used  as  fuel,  and  in  the  case  of  Pennsyl- 
vania oil,  distilled  under  vacuum  or  ordinary  pressure 
for  lubricating  oils,  the  residue  in  the  latter  case  being 


Petroleum  and  its  Products      155 

worked  up  for  vaseline,  solid  paraffins  or  paving  asphalt. 
If  the  distillation  is  carried  to  a  finish,  the  residue 
consists  of  a  coke-like  material  which  may  be  used  as  fuel. 

The  method  of  distillation  depends  on  the  nature  of  the 
petroleum  and  the  nature  of  the  products  required. 
Russian  petroleum  is  distilled  by  the  continuous  process, 
the  material  passing  through  a  series  of  stills  kept  at 
progressively  increasing  temperatures  ;  the  distillate  is 
divided  into  two  portions,  gasolene  and  kerosene.  The 
former  is  redistilled  till  the  distillate  has  a  specific 
gravity  of  0-750,  and  the  residue  added  to  the  main 
kerosene  fraction,  which  is  distilled  till  the  specific 
gravity  of  the  distillate  reaches  about  0-825.  The  total 
residue,  which  has  a  specific  gravity  of  0-903,  if  distilled, 
will  yield  about  10  per  cent,  of  solar  oil,  of  specific 
gravity  0-806  and  flash  point  105°  F.,  as  determined  by 
the  Pensky-Martens  apparatus,  and  about  35  per  cent,  of 
lubricating  oils,  the  final  residue  being  used  as  fuel.  Com- 
pared with  Pennsylvania  petroleum,  Russian  petroleum 
yields  a  relatively  small  proportion  of  light  petroleum 
and  burning  oil,  the  yield  of  the  former,  which  includes 
benzine  of  specific  gravity  about  0-725,  and  heavy 
benzine  or  gasolene  of  specific  gravity  about  0-770,  being 
about  4  to  6  per  cent.,  and  of  burning  oils,  about  27  per 
cent,  of  kerosene  of  specific  gravity  about  0-822  and 
5  per  cent,  of  solar  oil  of  specific  gravity  about  0-86  to 
0-88.  At  Baku,  about  30  to  40  per  cent,  is  generally 
distilled  off,  the  residue,  known  as  "  astatki,"  being  used 
as  liquid  fuel. 

Pennsylvania  petroleum  cannot  be  distilled  by  the 
continuous  process,  owing  to  the  large  amount  of  burning 
oil  which  it  yields,  and  the  viscous  nature  of  the  residue. 
When  submitted  to  the  simple  distillation  process,  it  first 
yields  gaseous  paraffins  which  are  either  burnt  as  fuel  or 
condensed  by  artificial  cooling  ;  the  liquid  products  thus 
obtained,  i.e.,  cymogene,  b.p.  32°  F.,  and  rhigolene, 
b.p.  65°  F.,  are  used  in  surgery.1  The  light  petroleum 

1  Distinction  must  be  made  between  benzine,  which  is  obtained  from  petroleum 
or  mineral  naphtha,  and  benzene,  the  aromatic  hydrocarbon  obtained  from  benzol, 
the  distillation  product  of  coal  tar  naphtha.  The  terms  petroleum  spirit  and 
light  petroleum  are  also  used  to  designate  the  more  volatile  distillation  products 
of  petroleum  naphtha. 


156      Industrial  Organic  Analysis 

or  petroleum  naphtha  fraction  is  taken  till  the  specific 
gravity  of  the  distillate  reaches  0-725  to  0750  ;  it 
includes  petroleum  ether,  specific  gravity  0-625,  gasolene, 
specific  gravity  0-665,  naphtha,  specific  gravity  0-676, 
and  benzine,  specific  gravity  0-736,  amounting  to 
9  to  18  per  cent,  of  the  total.  The  kerosene  fraction  is 
taken  until  the  specific  gravity  of  the  distillate  reaches 
0-840  to  0-845,  the  average  yield  of  burning  oils  being 
about  55  per  cent.  The  residue  is  transferred  to  other 
stills  and  distilled  for  lubricating  oils,  the  average  yield 
of  the  latter,  if  distilled  under  ordinary  pressure,  being 
about  17  per  cent.  The  solid  paraffin  amounts  to  about 
2  per  cent.,  and  coke  residue  about  10  per  cent.  Distilla- 
tion of  the  residue  by  the  vacuum  process  gives  a  much 
larger  yield  of  lubricating  oils  and  vaseline.  Sometimes 
the  heavier  oils  are  "  cracked  "  after  removal  of  the 
naphtha.  In  this  process  the  vapours  are  allowed  to 
condense  on  the  walls  of  the  still,  and  on  falling  back  into 
the  hot  residue,  decomposed  with  the  formation  of  lower 
boiling'  substances.  The  yield  of  burning  oils  may  thus 
be  increased,  it  being  possible  to  obtain  about  75  per  cent, 
of  oil  flashing  at  20°  C.,  as  indicated  by  the  Abel  appa- 
ratus, though  the  yield  of  higher  class  oils  of  higher  flash 
point  is  smaller. 

Regarding  the  petroleums  from  other  parts  of  the 
world,  the  Galician  and  Roumanian  oils  are  intermediate 
between  the  Pennsylvania  oils  on  the  one  hand,  and  the 
Russian  oils  on  the  other,  in  respect  to  chemical  com- 
position and  general  properties.  The  specific  gravity  of 
these  oils  is  usually  about  0-870.  The  petroleums 
obtained  from  Germany  and  Ohio,  U.S.A.,  also  contain 
both  paraffins  and  naphthenes.  Californian  petroleum 
contains  only  a  small  percentage  of  low  boiling  con- 
stituents ;  it  contains  no  paraffins,  but  naphthenes  and 
aromatic  hydrocarbons,  the  latter  often  being  present 
in  considerable  amount.  Appreciable  quantities  of 
oxygen  and  nitrogen  compounds  are  also  present. 
Texan  petroleum  does  not,  as  a  rule,  commence  to  distil 
below  240°,  and  consists  chiefly  of  unsaturated  hydro- 
carbons of  high  molecular  weight.  Canadian  petroleum 
contains  paraffins  and  olefines,  more  aromatic  hydro- 


Petroleum  and  its  Products      157 

carbons  and  less  light  paraffins  than  Pennsylvania 
oil. 

The  greater  part  of  the  world's  supply  of  burning  oil 
and  light  petroleum,  or  benzine,  is  obtained  from  the 
American,  and  some  of  the  Roumanian  and  Galician  oils. 
The  lubricating  oils  obtained  from  these  are  generally 
heavier,  more  viscous  and  of  higher  solidifying  point 
than  those  obtained  from  the  Russian  oils.  Oils  for 
lubricating  steam  cylinders,  which  require  to  be  fairly 
thick  and  viscous  on  account  of  the  relatively  high 
temperatures  at  which  they  are  used,  are  therefore 
obtained  best  from  the  American  petroleums.  Crude 
petroleums  containing  little  or  no  volatile  hydrocarbons, 
coming  under  the  heading  of  light  petroleum  or  burning 
oils,  will,  no  doubt,  find  increasing  use  as  liquid  fuel  for 
internal  combustion  engines  of  the  Diesel  type,  the  only 
preparation  necessary  being  the  removal  of  water  and 
solid  matter  by  the  settling  out  process  to  which  crude 
oils  are  generally  subjected.  The  use  of  Russian 
petroleum  residue  as  liquid  fuel  has  already  been 
alluded  to. 

The  various  fractions  obtained  by  the  distillation  of 
crude  petroleum  are  refined,  where  necessary,  by  treat- 
ment with  limited  quantities  of  concentrated  sulphuric 
acid,  and  subsequent  washing  with  dilute  caustic  soda 
solution  and  water.  Impurities  which  colour  the  oil  and 
give  it  an  objectionable  odour  are  thus  removed.  If 
sulphur  compounds  are  present  in  appreciable  quantity, 
they  are  removed  generally  by  treatment  with  lead  or 
copper  oxide  ;  either  the  fractions  are  redistilled  over 
the  metallic  oxide,  or  the  vapours  are  allowed  to  pass 
over  the  latter.  In  Canada,  burning  oils  are  desul- 
phurised by  treatment  with  a  solution  of  litharge  in 
caustic  soda  solution. 

THE  EXAMINATION  OF  CRUDE  PETROLEUM. 

Owing  to  the  complex  nature  of  crude  petroleums, 
these  are  but  rarely  examined  by  purely  chemical 
methods  in  technical  laboratories.  By  means  of  the 
distillation  test,  described  below,  the  nature  and 


158      Industrial  Organic  Analysis 


amount  of  the  distillate  to  be  expected  may  be  deter- 
mined, and  the  oil  may  be  roughly  classified.  The  specific 
gravity  may  also  be  used  to  help  in  classifying  and  com- 
paring petroleums.  The  determination  of  extraneous 
matter,  i.e.,  water  and  sediment,  will  also  be  described. 

Water. — A  weighed  quantity  of  the  oil,  about  100  to 

.  200    grams,    is   mixed   with    about    a 

third  of  its  bulk  of  xylene,  and  dis- 
tilled from  a  flask  heated  in  an  oil  bath ; 
the  distillate  is  collected  in  a  graduated 
cylinder  or  vessel  of  the  form  shown  in 
the  accompanying  illustration.  The  dis- 
tillation is  stopped  when  no  more  water 
is  seen  to  come  over,  and  the  vessel 
containing  the  distillate  is  warmed  and 
gently  shaken  so  that  the  water  adher- 
/  ing  to  the  sides  falls  to  the  bottom. 

The  volume  of  the  layer  of  water  at 
the  bottom  of  the  cylinder  or  in  the 
constricted  part  of  the  special  measur- 
ing vessel  is  read  off  after  cooling. 

Sediment. — 5  to  10  grams  of  the  oil 
are  well  mixed  with  100  to  500  grams 
of  benzene  and  allowed  to  stand  over- 
night ;  the  benzene  and  petroleum 
mixture  is  then  filtered  through  a 
weighed  filter,  and  the  insoluble  residue  washed  with 
benzene,  dried  at  100°  and  weighed.  The  sediment 
estimated  will  consist  of  mineral  matter,  and  will  not 
include  any  solid  pitch  or  asphalt  which  may  exist  in 
the  oil,  these  being  soluble  in  benzene. 

Specific  Gravity. — If  the  oil  is  not  too  thick,  a  hydro- 
meter may  be  used.  The  oil  is  poured  into  a  cylinder  of 
suitable  size  and  allowed  to  stand  for  some  time  to 


FIG.  9. — Measur- 
ing Vessel  for 
Water  in 
Petroleum. 


Petroleum  and  its  Products      159 

acquire  the  room  temperature.     The  float  is  allowed  to 
sink   gradually  into   the  oil,    and   its  level   read   after 

15  minutes,  the  reading  corresponding  with  the  level  of 
the  liquid  being  taken.     In  the  case  of  dark  oils,  the 
level  to  which  the  oil  rises  up  the  stem  of  the  float  is 
read  off,  and  0-0015  or  o-ooio  is  added  to  the  reading, 
according  as  the  paper  scale  is  shorter  or  longer  than 

1 6  cm.     The  temperature  may  be  read  off  from  the 
thermometer  on  the  float,   or,   failing   this,   the  room 
temperature  may  be  taken.     In    order   to   reduce   the 
reading    to    15°,   a   correction   of    0-00065   Per   degree 
Centigrade  may  be  applied. 

If  the  specific  gravity  bottle  is  used,  and  the  oil  is 
thick,  the  bottle  should  first  be  nearly  filled  with  the 
oil  and  allowed  to  stand  in  a  warm  place  overnight,  in 
order  that  all  air  bubbles  may  rise  to  the  top.  The 
bottle  and  its  contents  are  then  cooled  to  15°  by  allow- 
ing to  stand  in  a  water  bath  at  that  temperature, 
filled  up  and  weighed  in  the  usual  way. 

Distillation  Test. — The  test  to  be  described  is  originally 
due  to  Engler,  certain  modifications  having  been  intro- 
duced by  Ubbelohde.  The  method  is  empirical,  but  gives 
uniform  results  if  the  prescribed  dimensions  of  the 
apparatus  and  conditions  of  distillation  are  adhered  to. 
If  the  oil  contains  much  water,  it  should  first  be  dehy- 
drated by  means  of  calcium  chloride,  to  avoid  frothing 
in  the  distillation.  100  c.c.  of  the  oil  are  distilled 
from  a  flask  of  the  form  and  dimensions  shown  in 
Fig.  10.  A  Liebig  condenser  may  be  used,  and  the 
fractions  are  collected  so  that  they  may  be  weighed 
or  measured.  The  distillation  should  be  carried 
out  so  that  the  distillate  collects  at  a  uniform  rate 
of  two  drops  per  second.  At  first  the  flask  should  be 
heated  on  wire  gauze,  and  subsequently  a  free  flame 


160      Industrial  Organic  Analysis 

may  be  used.  The  first,  or  naphtha  fraction,  is  collected 
until  the  temperature  indicated  by  the  thermometer  is 
150°  ;  the  receiver  is  then  changed  for  the  kerosene 


fc 6:5 

FIG.  10. — Petroleum  Distillation  Flask  according  to  Engler. 

fraction,  which  is  collected  between  150°  and  300°. 
In  the  case  of  Russian  and  Galician  oils,  the  kerosene 
fraction  should  only  be  collected  up  to  285°  and  275° 
respectively,  the  oils  obtained  at  higher  temperatures 
from  these  petroleums  being  unsuitable  for  burning  in 


Petroleum  and  its  Products      161 

lamps  as  they  cause  excessive  charring  of  the  wick. 
What  remains  in  the  flask  after  distilling  off  the  kerosene 
is  weighed  as  residuum.  If  desired,  separate  fractions 
may  be  taken  at  more  frequent  intervals,  say,  every  25°  ; 
this  is,  however,  only  advisable  if  the  yield  of  distillate 
is  large.  If  the  oil  only  yields  very  little  distillate  up 
to  300°,  a  larger  flask,  say  of  140  c.c.  capacity,  must  be 
used,  or  the  quantity  of  oil  taken  must  be  reduced  to  80 
or  90  c.c. ;  the  heated  oil  will  otherwise  expand  so  as  to 
fill  the  flask  to  such  an  extent  that  it  may  be  carried 
over  mechanically  during  the  distillation.  The  same 
precautions  must  be  taken  if  the  material  under  examina- 
tion consists  of  petroleum  residue. 

Various  petroleums  of  known  origin  should  be  com- 
pared by  this  method.  Petroleum  is  examined  for 
excise  purposes  in  specially  constructed  metal  stills  of 
standard  dimensions.  For  a  method  of  examining 
petroleum  for  technical  purposes,  duplicating  the  manu- 
facturing operations  on  a  small  scale,  see  Holde's 
"  Untersuchung  der  Mineralole  und  Fette,"  p.  10. 

THE  EXAMINATION  OF  PETROLEUM  NAPHTHA  AND  ITS 
DISTILLATION  PRODUCTS. 

Petroleum  naphtha  is  usually  understood  to  include 
the  portion  of  the  distillate  from  crude  petroleum  which 
comes  over  below  150°  ;  on  redistillation,  however, 
it  is  generally  found  to  contain  up  to  10  per  cent,  of  oils 
distilling  above  this  temperature. 

The  various  products  which  are  obtained  by  the 
fractional  distillation  of  petroleum  naphtha  are  best 
characterised  by  their  specific  gravity,  which  may  be 
determined  by  means  of  a  float  or  the  Westphal  balance 
which  is  described  in  Chapter  II.,  p.  42.  The  results 

I.O.A.  II 


1 62      Industrial  Organic  Analysis 

may  be  calculated  to  15°  C.  from  the  data  given  in  the 
following  table,  which  is  due  to  Mendeleeff  :— 

For  Specific  Gravities  from  Correction  per  °  C. 

0700  to  0720         . .  . .         0-000820 

0720   ,,   0740  .  .  .  .  0-OOO8l8 

0-740   ,,   0-760  .  .  .  .  O-OOO80O 

0-760   ,,   0-780  .  .  .  .  0-000790 

0-780   ,,   0-8OO  .  .  .  .  0-000780 

The  most  important  uses  of  the  petroleum  "naphtha 
distillates  are  as  fuel  in  internal  combustion  engines,  and 
as  solvents  for  extraction  of  organic  material,  as,  for 
example,  the  extraction  of  fatty  oils  and  fats  from  seeds 
and  other  vegetable  products.  The  requirements  for 
material  which  is  to  be  used  for  these  purposes  are  as 
follows  : — 

It  should  be  water  white  and  free  from  objectionable 
smell ;  what  smell  it  possesses  should  be  as  faint  as 
possible.  An  objectionable  smell  is  sometimes  masked 
by  the  addition  of  small  quantities  of  turpentine  or  rosin 
oil  and  treatment  with  alkali  ;  such  additions  may  be 
detected  as  described  below.  The  specific  gravities  of  the 
better  qualities  of  motor  petrol  range  from  0-710  to  0-720, 
and  of  the  cheaper  qualities  from  0-728  to  about  0-735  or 
even  higher.  When  gently  heated  on  the  water  bath  it 
should  leave  no  oily  residue,  and  when  allowed  to 
evaporate  on  filter  paper  it  should  leave  no  mark.  A 
more  accurate  test  for  the  presence  of  heavier  oils  is  the 
distillation  test  described  below.  When  agitated  with 
water  the  latter  should  remain  neutral  and  give  no 
precipitate  with  barium  chloride  solution  ;  otherwise, 
the  oil  will  contain  some  of  the  sulphuric  acid  which  has 
been  used  in  the  refining  process. 

The  detection  and  estimation  of  aromatic  hydro- 
carbons will  be  described  below  (p.  136).  For  the  deter- 


Petroleum  and  its  Products     163 

initiation  of  flash  point,  which  for  the  present  purpose  is 
a  test  of  secondary  importance,  see  the  works  mentioned 
in  the  list  at  the  end  of  this  chapter. 

Fractional  Distillation  Test. — The  material  is  distilled 
in  a  flask  of  about  200  c.c.  capacity,  heated  on  a  sand 
bath,  and  fitted  with  a  fractionating  column  which 
carries  a  thermometer  and  connects  with  a  straight  tube 
water  condenser.  The  test  may  be  carried  out  as  de- 
scribed in  Chapter  IL,  p.  43. 

According  to  Holde,  the  best  benzine  for  motor  engines 
should  be  completely  volatile  below  100°,  or  should,  at 
least,  only  yield  up  to  5  per  cent,  of  distillate  above  this 
temperature.  If  it  contains  a  greater  percentage  of 
constituents  boiling  above  100°,  it  will  not  evaporate 
with  sufficient  rapidity  on  entering  the  explosion 
chamber,  especially  in  cold  weather.  It  is  also  obvious 
that  benzine  which  is  to  be  used  for  extraction  purposes 
must  not  contain  an  unduly  large  proportion  of  higher 
boiling  constituents,  or  it  will  only  be  removed  with 
difficulty  from  the  extracted  material. 

Aromatic  Hydrocarbons. — According  to  Holde,  these 
may  be  detected  as  follows J :  A  small  quantity  of  the 
benzine  is  treated  with  a  little  powdered  asphalt  (suffi- 
cient to  cover  the  point  of  a  penknife),  which  has  been 
freed  from  mineral  matter  (by  solution  in  benzene  and 
filtration)  and  repeatedly  washed  with  benzine  of 
specific  gravity  0-70  to  071  in  order  to  free  it  from  its 
more  soluble  constituents.  The  benzine  containing  the 
asphalt  is  passed  through  a  small  filter  into  a  test 
tube  ;  if  the  filtrate  is  colourless,  benzene  hydrocarbons 
are  absent,  if  brown,  benzene  or  toluene  are  probably 
present. 

If  a  few  cubic  centimetres  of  light  petroleum  containing 

1  Compare  also  the  method  for  distinguishing  petroleum  spirit  and  benzol  given 
in  Chapter  II.,  p.  55. 

II — 2 


164      Industrial  Organic  Analysis 

benzene  or  toluene  be  shaken  in  a  test  tube  with  a 
mixture  of  concentrated  nitric  and  sulphuric  acids,  the 
mixture  will  sooner  or  later  become  warm,  while  the 
upper  layer  will  develop  a  yellow  coloration ;  on 
pouring  into  water,  there  will  be  a  separation  of  yellow, 
oily  drops,  having  the  characteristic  odour  of  nitro- 
benzene. 

The  presence  of  aromatic  hydrocarbons  will  raise  the 
specific  gravity  of  the  sample.  (See  reference  in  foot- 
note on  p.  163.) 

The  quantitative  estimation  of  aromatic  hydrocarbons 
in  benzine  may  be  accomplished  by  the  following  method, 
due  to  Kraemer  and  Bottcher,  which  is  based  on  the  fact 
that  the  aromatic  hydrocarbons  (as  well  as  the  ethylenes) 
are  absorbed  by  concentrated  sulphuric  acid,  while  the 
paraffins  are  unaffected.  The  acid  required  is  prepared 
by  mixing  80  parts  by  volume  of  concentrated  sulphuric 
acid  of  specific  gravity  1-84  with  20  parts  of  fuming 
sulphuric  acid.  25  c.c.  of  the  benzine  (or  petroleum)  are 
placed  in  a  strong  75  c.c.  flask  which  has  a  neck  50  cm. 
long,  graduated  in  tenths  of  a  cubic  centimetre l ;  25  c.c. 
of  the  acid  are  then  added,  and  the  whole  is  shaken 
vigorously  for  15  minutes.  After  30  minutes,  acid  is 
filled  into  the  flask  until  the  whole  of  the  upper  oily 
layer  comes  into  the  graduated  neck,  and  the  volume  is 
read  off  every  hour  until  no  increase  in  the  amount  of 
the  indifferent  hydrocarbons  takes  place.  The  difference 
between  the  original  volume  and  the  final  volume  of  the 
oil  is  the  amount  of  hydrocarbons  absorbed  by  the  acid. 
If  more  than  13  per  cent,  of  hydrocarbons  be  absorbed 
the  test  is  inaccurate. 

Turpentine  and  Rosin  Oils. — Holde  recommends  the 
following  test  for  small  quantities  of  turpentine  or  rosin 

i  A  glass  stoppered  graduated  cylinder  may  also  be  employed. 


Petroleum  and  its  Products      165 

oils,  either  in  pure  petroleum  benzine  or  in  petroleum 
benzine  containing  benzene  or  homologues  of  the  latter. 
Bromine  vapour  is  allowed  to  flow  into  a  test  tube  con- 
taining a  little  of  the  benzine  ;  on  shaking  the  benzine 
should  immediately  take  up  the  red  colour  of  the 
bromine  ;  if,  however,  turpentine  or  rosin  oils  be  present, 
traces  of  added  bromine  will  rapidly  be  absorbed,  owing 
to  the  unsaturated  nature  of  these  substances. 

THE  EXAMINATION  OF  KEROSENE  OR  BURNING  OIL. 

Kerosene,  or  oil  for  burning  in  wick  lamps,  is  obtained 
by  redistilling  the  fraction  which  succeeds  the  naphtha 
fraction  in  the  distillation  of  crude  petroleum.  The 
refining  process  usually  consists  in  treatment  with  strong 
sulphuric  acid  and  subsequent  washing  with  dilute 
alkali  and  water,  and,  where  necessary,  desulphurisation 
as  indicated  above.  The  specific  gravity  of  American 
kerosene  usually  lies  between  0-790  and  0-800,  and  that 
of  Russian  kerosene  between  0-821  and  0-823,  the  latter 
generally  being  a  more  homogeneous  product  than  the 
former.  A  parallel  product  obtained  from  shale  oil  has 
a  specific  gravity  of  about  0-800.  A  good  kerosene 
should  be  water  white,  or,  at  the  most,  of  a  light  yellow 
colour.  Although  the  degree  of  colour  of  burning  oil 
is  often  made  the  basis  of  commercial  transactions,  this 
cannot  be  regarded  as  giving  any  real  indication  as  to 
the  degree  of  refinement  or  the  ability  of  the  oil  to  burn 
free  from  soot  and  smell  and  with  a  steady  flame.1 
Burning  oils  are  usually  examined  colorimetrically  by 
means  of  Stammer's  colorimeter.  Instructions  for  the 
use  of  this  instrument  are  supplied  by  the  makers, 
Messrs.  Schmidt  an^i  Haensch  of  Berlin.  Various 
colorimeters  will  also  be  found  described  in  the  works 
mentioned  at  the  end  of  this  chapter. 

The  most  important  tests  to  be  carried  out  on  burning 

1  Vorschriften  der  Russischen  Regierung  nach  Einvernehmen  mit  der  kaiserl. 
Russische  Technische  Gesellschaft  in  Baku  betr.  Nomenklatur  und  Prtifung  von 
Russischen  Erdolproducten. 


i66      Industrial  Organic  Analysis 

oils  are  the  distillation  test,  the  specific  gravity,  the 
flash  point,  and  the  sulphuric  acid  test  for  the  degree  of 
refinement.  Impurities  such  as  sulphur  and  free  acid 
should  also  be  tested  for. 

Distillation  Test. — This  test  may  be  carried  out  with 
the  Engler  apparatus,  as  described  above  for  crude 
petroleum  (p.  159).  The  temperature  of  the  beginning 
of  the  distillation  is  taken  when  the  first  drop  of  dis- 
tillate falls  from  the  end  of  the  condenser.  Fractions 
are  taken  every  50°  from  150°  to  250°,  and'  then 
every  25°  up  to  300°.  What  remains  is  estimated  by 
difference.  The  end  point  for  each  fraction  is  the  point 
at  which  after  cooling  at  least  20°  and  reheating  to  the 
highest  temperature  at  which  the  fraction  is  to  be  taken, 
not  more  than  6  drops  of  distillate  run  from  the  con- 
denser, the  process  of  cooling  and  heating  being  repeated 
till  this  result  is  obtained.  The  volume  of  each  fraction 
is  measured  after  cooling  to  the  room  temperature. 

Distillation  should  not  commence  below  110°,  the 
yield  below  150°  should  be  at  most  10  per  cent.,  and  the 
yield  above  300°  at  most  15  per  cent.  Better  class  oils 
yield  85  to  go  per  cent,  between  150°  and  300°,  and,  at 
most,  5  per  cent,  above  300°.  The  higher  the  per- 
centage of  light  oils,  the  greater  the  danger  of  explosion 
when  the  oil  is  burnt  in  a  lamp,  while  the  higher  the 
percentage  of  high  boiling  constituents,  the  more  liable 
will  the  oil  be  to  clog  in  the  wick  and  cause  charring  and 
uneven  burning. 

Specific  Gravity. — The  specific  gravity  of  kerosene  may 
be  determined  by  any  of  the  usual  methods,  as  indicated 
under  the  heading  of  petroleum  naphtha,  The  tempera- 
ture correction  per  °  C.  for  reducing  results  to  15°,  as 
determined  by  Mendeleeff,  varies  from  0*000790  for 
specific  gravities  between  0760  and  0780,  to  0-000710 
for  specific  gravities  between  0-850  and  0-860.  The 


Petroleum  and  its  Products      167 

specific  gravity  is  mainly  used  as  a  test  of  identity,  and 
is  of  little  use,  taken  by  itself ;  thus  a  burning  oil  might 
very  well  contain  large  proportions  both  of  light  and 
heavy  oils  and  yet  show  a  normal  specific  gravity  ;  in 
a  case  like  this,  the  inferior  quality  of  the  oil  would  only 
be  revealed  by  the  distillation  test. 

Flash  Point. — The  flash  point  of  burning  oils  is  usually 
determined  in  specially  constructed  "closed  testers," 
which  will  be  found  described  in  the  works  of  reference 
mentioned  at  the  end  of  this  chapter.  The  oil  is 
gradually  heated  while  a  small  flame  is  brought  near 
its  surface  from  time  to  time  ;  the  temperature  at 
which  the  oil  begins  to  give  off  sufficient  inflammable 
vapour  to  produce  a  flash  is  taken  as  the  flash  point. 
The  test  thus  gives  an  indication  of  the  degree  of  safety 
with  which  the  oil  may  be  used  in  a  lamp  ;  the  greater 
the  proportion  of  lighter  constituents,  as  shown  by  the 
distillation  test,  the  lower  will  be  the  flash  point  and  the 
greater  the  danger  of  explosion.  In  Great  Britain  and 
Canada  the  Abel  tester  has  been  adopted  as  the  standard 
instrument,  the  minimum  flash  points,  as  laid  down  by 
the  laws  of  these  countries,  being  73°  F.  and  85°  F., 
respectively.  In  Germany  and  Russia,  both  the  Abel 
and  the  Pensky-Martens  instruments  are  used,  the 
minimum  flash  point  allowable  being  21°  C.  (70°  F.)  in 
Germany,  and  28°  C.  (84-4°  F.)  in  Russia.  In  the 
United  States  various  standard  instruments  and  minima 
have  been  adopted  in  different  states. 

Degree  of  Refinement. — The  following  method  for 
determining  the  degree  of  refinement  of  burning  oils  by 
the  colour  which  they  impart  to  concentrated  sulphuric 
acid  has  been  recommended  by  the  Baku  section  of  the 
Russian  Technical  Society.  100  parts  by  volume  of  the 
oil  are  shaken  for  2  minutes  in  a  glass-stoppered  cylinder, 
with  40  parts  by  volume  of  sulphuric  acid  of  specific 
gravity  1-73,  at  a  temperature  not  exceeding  32°. 
The  acid  is  then  separated  off  and  transferred  to  a 


1 68      Industrial  Organic  Analysis 

cylinder  of  white  glass,  where  it  may  be  compared,  as 
regards  colour,  with  solutions  of  Bismark  brown,  of 
known  strengths,  contained  in  similar  cylinders.  The 
layers  of  liquid  observed  should  be  of  equal  depth,  and 
the  cylinders  should  be  placed  on  a  uniform  white  sur- 
face. The  test  solutions  are  prepared  as  follows  :  a 
solution  containing  0-5  gram  of  the  colour  per  litre  is 
first  prepared,  and  from  this  ten  solutions  are  made  by 
dilution  with  water  ;  the  first  and  lightest  contains  i 
part  of  the  stock  solution  to  99  parts  of  water,  the 
second,  2  parts  to  98  parts  of  water,  and  so  on  to  the 
tenth  and  darkest,  which  contains  10  parts  of  the  stock 
solution  to  90  parts  of  water.  These  solutions  are 
numbered  from  I  to  10,  passing  from  the  lightest  to  the 
darkest,  and  the  burning  oil  is  marked  according  to  the 
solution  with  which  the  sulphuric  acid  extract  most 
nearly  corresponds  in  tint. 

As  most  petroleums  are  found  to  come  within  the  limits 
of  i  to  8,  the  latter  mark  is  taken  to  be  the  maximum 
limit  for  a  marketable  product. 

Impurities. — Certain  petroleums,  notably  those  from 
Ohio  and  Canada,  contain  notable  amounts  of  sulphur 
compounds  which  it  is  necessary  to  remove  by  special 
processes  of  refining,  as  kerosenes  containing  more  than 
the  slightest  traces  of  such  impurities  will  give  a  notice- 
able odour  of  sulphur-  dioxide  on  burning.  Owing  to  the 
relatively  small  amounts  of  sulphur  present,  even  in  the 
worst  cases,  special  methods  must  be  employed  for  the 
determination  of  this  constituent.  One  of  the  methods 
most  commonly  employed  is  that  of  Allen,  modified  by 
Heussler  and  Engler,  in  which  a  known  weight  of  the 
petroleum  is  burnt  in  a  specially  constructed  lamp,  and 
the  products  of  combustion  led  over  a  solution  of 
potassium  hypobromite,  which  is  distributed  over  glass 
beads.  The  sulphur  is  then  estimated  as  sulphate  in 
this  solution.  The  above  method  is  described  in  most 
of  the  works  of  reference  mentioned  at  the  end  of  this 


Petroleum  and  its  Products      169 

chapter.     According  to  Holde,  a  good  burning  oil  should 
not  contain  more  than  0-02  per  cent,  of  sulphur. 

Burning  oils  should  be  free  from  acid  ;  about  20  grams 
of  the  oil  dissolved  in  a  neutral  mixture  of  ether  and 
alcohol  containing  a  little  phenol  phthalein  should  show 
a  permanent  pink  coloration  on  the  addition  of  one  drop 
of  decinormal  sodium  hydroxide  solution. 

Ash  may  be  determined,  according  to  Holde,  by  the 
following  method  :  500  c.c.  or  a  litre  of  the  burning  oil  is 
distilled  from  a  retort  until  10  c.c.  are  left  as  still  residue  ; 
the  latter  is  transferred  to  a  platinum  dish,  by  the  use  of 
light  petroleum,  evaporated  to  dryness,  and  the  residue 
incinerated. 

Good  burning  oils  should,  according  to  Holde,  contain 
not  more  than  2  milligrams  of  ash  per  litre. 

THE  EXAMINATION  OF  LUBRICATING  OILS. 

Most  of  the  lubricating  oils  used  nowadays  are  obtained 
from  the  residue  which  remains,  after  distilling  off  the 
naphtha  and  burning  oil  from  crude  petroleum.  This 
residue,  which  by  itself  boils  above  300°,  is  distilled  in 
a  current  of  superheated  steam  at  temperatures  varying 
from  180°  to  250°,  whereby  various  fractions  are 
obtained  which  show  considerable  differences  in  viscosity 
and  flash  point.  Deodorisation  is  effected  by  blowing 
with  air  and  treatment  with  strong  sulphuric  acid,  the 
latter  process  being,  as  usual,  followed  by  washing  with 
caustic  soda  solution  and  water.  Clarification  is 
effected  by  filtering  over  fuller's  earth  or  charcoal.  The 
lighter  lubricating  oils,  which  vary  in  colour  from  light 
yellow  to  brown,  and  have  specific  gravities  ranging 
from  0-895  to  0-900,  are  invariably  obtained  by  distilla- 
tion ;  the  products  of  higher  specific  gravities,  up  to 
about  0-940,  may  consist  either  of  "  reduced  oils,"  i.e., 
filtered  residues  from  which  the  more  volatile  portions 
have  been  removed  by  distillation,  or  of  oils  which  have 
been  distilled  in  vacua  or  in  a  current  of  superheated 


i yo      Industrial  Organic  Analysis 

steam.     Undistilled  products  are  generally  used  in  the 
cheaper  grades  of  lubricating  oils. 

Certain  non-drying  or  semi-drying  fatty  oils  are  used 
as  lubricants  for  delicate  machinery,  as,  for  example, 
neat's  foot  oil  and  spermaceti  oil,  which  are  used  for 
watches  and  clocks.  Generally,  however,  when  fatty 
oils  are  used  at  all,  they  are  used  in  admixture  with 
mineral  oils.  Such  blended  oils  generally  contain  rape 
or  cotton  seed  oils  which  have  sometimes  been  blown 
with  air  at  elevated  temperatures  (see  Chapter  III.,  p.  95 ) 
in  order  to  increase  their  viscosity.  The  detection  and 
estimation  of  fatty  oils  in  presence  of  mineral  oils,  and 
the  reasons  for  their  use,  will  be  treated  of  below.  For 
the  present,  it  may  be  mentioned  that  the  tendency  for 
fatty  oils  to  undergo  hydrolysis  and  decomposition 
under  the  action  of  steam  is  greatly  reduced  when  they 
are  mixed  with  mineral  oils.  Rosin,  rosin  oil,  coal  tar 
or  lignite  oils  must  be  considered  as  adulterants  if  found 
in  lubricating  oils  purporting  to  be  of  superior  grade. 
The  detection  of  such  constituents  will  be  described 
below. 

The  requirements  of  a  lubricating  oil  will  naturally 
vary  to  a  great  extent  with  the  conditions  under  which 
it  is  to  be  used.  In  general,  howeyer,  it  may  be  stated 
that  lubricating  oils  should  be  free  from  acids,  either 
mineral  or  organic,  which  may  cause  corrosion  of  the 
bearings,  and  also  from  all  materials  which  may  give 
rise  to  the  "  gumming  "  of  the  oil  when  spread  in  a  thin 
layer,  such  as  rosin,  rosin  oil  or  drying  fatty  oils.  Water 
or  sediment  of  any  kind  should  be  absent. 

The  function  of  a  lubricating  oil  is  to  keep  two  metallic 
surfaces  from  actual  contact  with  one  another  ;  for  this 
it  is  necessary  that  it  should  possess  a  certain  "  body  "  or 
viscosity  ;  the  greater  the  pressure  between  the  surfaces, 
the  greater  the  viscosity  required.  On  the  other  hand, 
the  higher  the  viscosity  of  the  oil,  the  greater  its  internal 
friction  ;  it  is,  therefore,  always  advisable  to  use  an  oil 
of  the  minimum  permissible  viscosity  in  order  to  avoid 
excessive  heating  of  the  bearings.  The  temperature  at 
which  the  oil  is  to  be  used  must  also  be  taken  into  con- 
sideration ;  thus,  the  oils  employed  for  the  lubrication  of 


Petroleum  and  its  Products      171 

steam  cylinders  are  of  a  thick  and  syrupy  consistency 
at  the  ordinary  temperature,  as  the  viscosity  invariably 
decreases    with    rise    of    temperature.     Fatty    oils    are 
sometimes  used  in  cylinder  oils,  as  they  lose  viscosity 
with  rise  in  temperature  to  a  lesser  degree  than  do  the 
mineral  oils.     Oils  which  are  to  be  used  in  refrigerating 
machines,    or    compressors,    are    quite    mobile    at    the 
ordinary  temperature,  as  the  viscosity  will  increase  on 
cooling.     The  viscosity  of  a  lubricating  oil,  as  deter- 
mined by  means  of  a  viscosimeter,  gives  an  indication  of 
the  class  of  work  for  which  it  is  best  suited,  though  in 
this  case,  practical  experience  should  be  the  main  guide. 
The  Engler  and  Redwood  viscosimeters  are  the  standard 
instruments  adopted  for  the  testing  of  lubricating  oils 
in   Germany   and    the   United   Kingdom    respectively. 
The    construction    and    use    of    these    instruments    is 
described   in  the  works  mentioned  at  the  end  of  this 
chapter.      Various  machines  for  testing  the  lubricating 
value  of  oils  wijl  also  be  found  described  in  these  works  ; 
it  is,  however,  pointed  out  by  Holde  that  these  machines 
often  fail  to  reproduce  the  conditions  occurring  in  actual 
practice. 

Other  physical  properties  which  must  be  taken  into 
account  are  the  setting  point  and  the  flash  point.  The 
determination  of  these,  and  their  significance  in  special 
cases,  will  be  dealt  with  below. 

Lubricating  oils  are  classified  by  Holde  ("  Untersuchung 
der  Mineralole  und  Fette  ")  as  follows  : — 

(1)  Spindle  Oils,  for  Textile  Machinery.     Mobile  oils  to 
be   used   under   light   pressures ;     viscosity   in   Engler 
degrees  at  20°  C.,  5  to  12 ;  flash  point  (Pensky),  160°  to 
200°  C. 

(2)  Compressor  Oils,  or  oils  for  use  in  refrigerating 
machines.     Mobile  oils  having  low  viscosities  ;    Engler 
degrees  at  20°  C.,  5  to  7.     It  is  important  that  these  oils 
when  submitted  to  the  cold  test  should  remain  liquid  at 

-  20°.     The  flash  point  may  be  low  (Pensky),  170°  to 
180°.   Compressor  oils  are  often  coloured  a  reddish  violet. 

(3)  Light  Engine  Oils,  suitable  for  motors  or  dynamos. 
These  are  moderately  viscous  ;  Engler  degrees  at  20°  C., 
13  to  25  ;  flash  point  (Pensky),  170°  to  220°. 


172      Industrial  Organic  Analysis 

(4)  Heavy  Engine  Oils.     Engler  degrees  at  20°  C.,  25 
to  45,  or  in  special  cases  up  to  60  ;  flash  point  (Pensky), 
190°  to  220°. 

The  foregoing  oils,  when  viewed  in  a  test  tube,  are 
light  brownish  yellow  to  brownish  red  in  colour,  the 
more  expensive  grades  being  light  yellow.  The  following 
varieties  are  dark  in  colour  and  often  opaque. 

(5)  Dark   Railroad    Oils.     These    are    classified    into 
summer  and  winter  oils.     The  former  show  viscosities 
of  45°  to  60°  Engler  at  20°  C.,  and  the  latter  25°  to  45°. 
The  flash  point  (Pensky)  should  be  over  140°.     Setting 
point,   for  summer   oils,  under   —  5°,   for  winter  oils, 
under  —  20°.     (The  above  requirements  for  railroad  oils 
are  based  on  the  climatic  conditions  of  Germany  ;    in 
England  it  is  not  necessary  to  prescribe  so  low  a  cold 
test  in  winter.) 

(6)  Cylinder  Oils.     These  are  of  a  syrupy  or  sometimes 
even  pasty  consistency  at  the  ordinary  temperature. 
The  viscosity  at  50°  C.  is  23°  to  45°  Engler.      Oils  for 
superheated  steam  cylinders  have  still  higher  viscosities, 
i.e.,  50°  to  60°  Engler  at  50°  C.     The  flash  point  should 
be   over   200°   and,    according    to   the   quality,   varies 
from   220°   to  315°  ;    the   better  qualities   flash   above 
260°.    Low  volatility  is  also  a  desideratum  for  cylinder 
oils  ;    this  generally  follows  with  a  high  flash  point. 
Cylinder  oils  which  have  been  filtered  over  fuller's  earth 
are   of   a  brownish  red  colour  and  transparent ;    the 
unfiltered  oils  are  greenish  black  and  opaque. 

Flash  Point. — The  determination  of  flash  point  by  the 
open  test  is  carried  out  as  follows  :  about  50  c.c.  of  the 
oil  to  be  tested  are  placed  in  a  porcelain  or  nickel  crucible 
of  about  75  c.c.  capacity.  The  crucible  is  placed  in  a 
tin  of  convenient  size  to  protect  the  surface  of  the  oil 
from  draughts,  the  bottom  of  which  is  covered  with 
about  half  an  inch  of  sand.  The  tin  is  heated  from 
below,  and  the  temperature  of  the  oil  is  indicated  by  a 
thermometer,  the  bulb  of  which  is  immersed  in  the  oil 
without  touching  the  bottom  of  the  crucible.  When  the 


Petroleum  and  its  Products      173 

temperature  has  been  raised  to  about  120°,  the  oil  is 
tested  from  time  to  time  by  bringing  a  small  gas  jet, 
burning  from  a  hard  glass  capillary  tube,  near  its  surface. 
The  lowest  temperature  at  which  the  vapour  from  the 
oil  is  observed  to  ignite  is  taken  as  the  flash  point. 
When  a  flash  has  been  obtained,  the  oil  may  be  cooled 
about  10°  and  reheated  at  the  rate  of  about  2°  per 
minute  to  get  a  more  accurate  determination  of  the 
flash  point.  It  should,  however,  be  noted  that  the  flash 
point  will  be  sensibly  raised  if  the  oil  is  kept  too  long  at 
an  elevated  temperature.  The  results  obtained  by  the 
open  test  are  not  so  accurate  as  those  obtained  by  means 
of  the  closed  testers,  e.g.,  the  Abel  or  the  Pensky-Martens 
instruments,  but  for  practical  purposes,  it  is  only  neces- 
sary to  be  able  to  say  whether  the  oil  will  be  dangerous 
or  not.  The  flash  points  indicated  by  the  closed  testers 
are  usually  about  15°  lower  than  those  obtained  by 
the  open  test. 

The  determination  of  flash  point  is  of  especial  import- 
ance in  the  case  of  oils  which  are  to  be  used  for  textile 
machinery  and  oils  for  lubricating  steam  cylinders.  If 
the  oil  has  a  high  flash  point,  it  is  not  only  less  liable  to 
damage  the  india-rubber  packings  of  the  cylinders  by 
ignition,  but  also  less  liable  to  loss  by  evaporation  on 
prolonged  exposure  to  elevated  temperatures.  The 
lighter  oils  for  lubricating  bearings  and  journals  usually 
flash  between  175°  and  200°  by  the  open  test.  The 
heavier  machine  oils  should  show  flash  points  of  200°  and 
upwards,  while  cylinder  oils  of  good  quality  should  flash 
above  280°  by  the  open  test.  Considerable  quantities  of 
fatty  oils  are  sometimes  added  to  mineral  cylinder  oils 
of  inferior  quality  in  order  to  counterbalance  their 
tendency  to  evaporate  at  the  temperature  of  the  steam 
cylinder,  and  their  low  flash  points. 

Setting  Point. — The  following  method  for  the  deter- 
mination of  the  setting  point  of  lubricating  oils  is 


174      Industrial  Organic  Analysis 

recommended  by  the  Scottish  Mineral  Oil  Association. 
The  oil  is  cooled  in  a  large  strong  walled  test  tube  until 
it  completely  solidifies.  In  the  absence  of  liquid  carbon 
dioxide  or  air,  the  following  freezing  mixtures  may  be 
employed  :  a  mixture  of  2  parts  of  snow  or  pounded  ice 
with  i  part  of  sodium  chloride,  which  gives  a  minimum 
temperature  of  —  23°,  or  a  mixture  -of  12  parts  of  snow 
or  pounded  ice,  5  parts  of  sodium  chloride  and  5  parts 
of  ammonium  nitrate,  which  gives  a  minimum  tempera- 
ture of  —  30°.  When  the  oil  has  solidified,  the  test  tube 
is  removed  from  the  freezing  mixture  and  held  up  to 
the  light,  while  the  oil  is  stirred  with  a  thermometer  ; 
the  temperature  at  which  the  last  trace  of  paraffin  dis- 
appears is  taken  as  the  setting  point.  The  test  is 
repeated  until  two  determinations  give  concordant 
results. 

As  mentioned  above,  lubricating  oils  for  refrigerating 
machines  should  remain  liquid  at  —  20°.  The  setting 
point,  or,  as  it  is  sometimes  called,  the  cold  test,  is  also 
of  importance  in  the  case  of  lubricating  oils  for  machinery 
which  is  used  in  the  open  in  countries  where  the  winter 
is  severe  ;  thus  for  some  winter  oils  it  is  necessary  to 
prescribe  a  setting  point  of  —  15°  or  —  20°.  In  England, 
however,  the  cold  test  is  of  less  importance.  If  the  oil 
should  solidify  in  the  journals  or  bearings,  friction  would 
be  developed  with  consequent  damage  to  the  machinery. 

Specific  Gravity. — In  the  case  of  the  lighter  oils,  a 
float  may  be  used  ; .  for  thicker  oils,  it  will  be  necessary 
to  use  a  specific  gravity  bottle,  as  described  for  crude 
petroleum. 

The  specific  gravity  has  no  relation  to  the  lubricating 
power  of  an  oil,  but  mainly  serves  as  a  means  of  classi- 
fication and  identification.  The  specific  gravities  of 
pure  mineral  lubricating  oils  usually  lie  between  0-884 
and  0-930  at  15°.  The  specific  gravity  cannot  be  looked 


Petroleum  and  its  Products      175 

on  as  proportional  to  the  viscosity  of  the  oil,  but  it  may 
be  said  that,  as  a  rule,  oils  of  high  specific  gravities  are 
used  for  high  pressures. 

Water. — Ordinary  lubricating  oils  should  contain  no 
water.  In  light  coloured  oils  water  is  easily  detected 
by  inspection  on  shaking  the  sample  ;  in  dark  oils  it 
will  be  revealed  by  the  bumping  or  spitting  which  will 
take  place  on  heating.  If  desired,  any  water  which  may 
be  present  may  be  estimated  as  described  for  crude 
petroleum. 

Acidity. — The  free  acids  liable  to  occur  in  lubricating 
oils  may  consist  of  sulphuric  acid  which  has  remained  in 
the  oil  after  the  refining  process,  or  organic  acids  arising 
from  the  decomposition  of  the  mineral  or  vegetable  oils. 
A  high  acidity  may  also  be  due  to  the  presence  of  rosin  or 
rosin  oil,  the  detection  of  which  will  be  described  below. 

Sulphuric  acid  may  easily  be  detected  by  shaking  out 
the  oil  with  warm  water,  and  testing  the  latter  when 
separated,  with  barium  chloride  solution  and  hydro- 
chloric acid  ;  this  impurity  will,  however,  only  be  found 
in  very  exceptional  cases. 

The  acidity  of  lubricating  oils  may  be  determined  by 
titration  as  described  under  the  heading  "  Acid  Value  " 
in  Chapter  III.  If  difficulty  is  experienced  owing  to 
the  dark  colour  of  an  oil,  the  following  process,  recom- 
mended by  Holde,  may  be  employed  :  20  c.c.  of  the  oil 
are  shaken  in  a  glass  stoppered  measuring  cylinder  with 
40  c.c.  of  neutral  absolute  alcohol,  the  oil  being  warmed 
if  not  sufficiently  mobile.  The  cylinder  is  stoppered 
and  allowed  to  stand  overnight,  after  which  20  c.c. 
of  the  clear  alcoholic  layer  are  titrated  with  deci- 
normal  sodium  hydroxide  solution,  using  phenol  phtha- 
lein  as  indicator.  If  more  than  0-03  per  cent,  of  acid  (cal- 
culated to  SO3)  is  found,  the  rest  of  the  alcohol  should 
be  poured  off,  the  above  process  repeated  with  the 


176      Industrial  Organic  Analysis 

residual  oil  and  a  further  quantity  of  40  c.c.  of  neutral 
alcohol,  and  the  amount  of  acid  found  in  this  second 
titration  added  to  that  found  in  the  first. 

Refined,  clear  mineral  oils  should  contain  no  free  acid, 
or,  at  most,  only  0-03  per  cent.,  calculated  as  SO8. 
Dark  oils  should  not  contain  more  than  0-3  per  cent,  of 
acids  ;  generally,  however,  the  proportion  of  these  con- 
stituents will  be  under  0-15  per  cent.  If  the  amount  of 
free  acids  in  a  lubricating  oil  is  above  the  prescribed 
limits,  there  is  danger  of  the  bearings  on  which  it  is  used 
becoming  corroded- 

Soap  and  Ash. — The  better  class  lubricating  oils  will 
contain  no  soap  or  foreign  inorganic  matter,  as  will  be 
shown  by  their  complete  solubility  in  benzene.  Soap 
may  be  added  in  order  to  increase  the  consistency  of  the 
oil  or  to  facilitate  its  emulsification  with  water ;  it  will 
be  recognised  by  the  ease  with  which  emulsions  are 
formed,  and  the  feebly  alkaline  reaction  of  these  towards 
phenol  phthalein.  An  examination  of  the  aqueous 
extract  will  soon  show  whether  soap  is  present  (see 
Chapters  III.  and  IV.,  introductory  sections),  and  the 
nature  of  the  base  (usually  soda  potash,  lime  or  ammonia) 
may  be  ascertained  by  examining  the  acidified  aqueous 
extract.  The  soap  may  be  estimated  as  described  for 
lubricating  greases. 

Lubricating  Greases. — Soap  is  chiefly  used  in  lubrica- 
ting greases,  in  which  it  is  emulsified  with  mineral,  lignite, 
rosin  or  tar  oils,  with  the  addition  of  a  little  water.  The 
soap  may  be  estimated  by  carefully  burning  off  a  weighed 
quantity  of  the  sample,  incinerating  and  titrating  the 
residual  potassium  carbonate,  sodium  carbonate,  or  lime, 
with  semi-normal  hydrochloric  acid.  The  amount  of  the 
base  found  may  then  be  calculated  to  stearate  or  rosinate 
(See  Chapter  IV.,  p.  128.)  If  the  base  present  is  ammonia, 
this  must  be  estimated  by  one  of  the  usual  methods  in 


Petroleum  and  its  Products       177 

the  aqueous  acid  extract  obtained  from  a  known  weight 
of  the  sample  by  extracting  it  with  dilute  mineral  acid. 
Water  may  be  estimated  in  lubricating  greases  by  the 
xylene  distillation  method,  as  described  for  crude  petro- 
leum. The  oily  portion  may  be  separated  for  examina- 
tion for  the  constituents  mentioned  above  by  adding 
ether,  extracting  several  times  with  water  to  remove  the 
soap  and  then  evaporating  off  the  ether.  If  desired,  the 
aqueous  soap  solution,  which  has  been  entirely  freed 
from  oily  matter,  may  be  examined  as  described  in 
Chapter  IV. 

DETECTION  AND  ESTIMATION  OF  FATTY  OILS,  ROSIN, 
ROSIN  OIL,  TAR  OIL,  ETC.,  IN  PRESENCE  OF  MINERAL 

OILS. 

Fatty  Oils. — As  a  preliminary  test,  the  saponification 
value  of  the  sample  may  be  determined  as  described  in 
Chapter  III.,  p.  95.  If  no  definite  value  is  obtained,  then 
fatty  oils  may  be  assumed  to  be  absent.  The  following 
method  for  the  rapid  determination  of  fatty  oils  in 
presence  of  mineral  oils  is  recommended  by  Schreiber  :— 

5  grams  of  the  oil  are  weighed  out  in  a  200  c.c.  conical 
flask,  25  or  50  c.c.  of  approximately  semi-normal 
alcoholic  potash  (see  Chapter  III.,  p.  96)  are  added,  and 
then  sufficient  benzene  to  dissolve  the  oil  when  warmed. 
25  c.c.  of  benzene  will  generally  be  sufficient,  but  for 
cylinder  oils  50  c.c.  may  be  necessary  ;  in  this  case, 
25  c.c.  of  neutral  alcohol  may  be  added  with  advantage. 
At  the  same  time,  a  blank  experiment  is  started  with 
equal  amounts  of  alcoholic  potash,  benzene  and  alcohol, 
to  those  used  in  the  actual  determination.  The  flasks 
are  connected  with  air  condensers  and  placed  on  a  water 
bath  so  that  the  contents  are  kept  gently  simmering. 
During  the  saponification,  which  will  be  complete  in 

I.O.A.  12 


178      Industrial  Organic  Analysis 

30  minutes,  the  contents  of  the  flasks  should  be  gently 
shaken  from  time  to  time.  The  amount  of  potash  used 
up  in  the  saponification  is  determined  by  titration  with 
semi-normal  acid  in  presence  of  phenol  phthalein,  as 
described  under  the  heading  "  Saponification  Value  " 
in  Chapter  III.  For  practical  purposes,  the  saponi- 
fication value  of  the  fatty  oils  used  in  lubricating 
oils  may  be  taken  as  195.  Then,  if  S  be  the  saponifica- 
tion value  found,  the  percentage  of  fatty  oil  present  will 

,     100  S 
be  -    — • 
195 

Some  cylinder  oils  may  contain  about  30,  or  even 
60  per  cent,  of  fatty  oils.  The  reasons  for  such  additions 
have  already  been  explained. 

Rosin  (see  also  Chapter  IV.). — As  was  pointed  out 
above,  a  somewhat  high  acidity  may  be  due  to  the 
presence  of  dissolved  rosin.  In  light  coloured  mineral 
oils,  rosin  may  be  detected  by  the  Liebermann-Storch 
reaction,  as  described  in  Chapter  IV.,  p.  144.  If  the  oil 
be  too  dark,  the  rosin  acids  may  be  dissolved  out  by 
extracting  with  dilute  sodium  hydroxide  solution  and 
precipitated  by  acidifying  the  alkaline  extract.  This 
process  may  also  be  adapted  for  the  quantitative  estima- 
tion of  rosin,  provided  that  fatty  oils  are  not  present  ; 
the  precipitated  acids  are  filtered  off,  dried,  weighed, 
after  which  they  may  be  tested  by  the  Liebermann- 
Storch  reaction. 

A  quicker  way  of  separating  the  rosin  for  qualitative 
examination  is  to  shake  about  10  c.c.  of  the  oil  with  an 
equal  volume  of  hot  70  per  cent,  alcohol,  and  after 
cooling,  filtering  and  evaporating  the  alcoholic  layer. 
If  rosin  be  present,  the  residue  will  be  of  a  resinous 
nature,  and  not  oily,  and  it  will  give  the  characteristic 
coloration  in  the  Liebermann-Storch  test. 


Petroleum  and  its  Products       179 

If  fatty  oils  are  present,  it  will  be  necessary  to 
saponify  as  described  under  the  heading  "  Fatty  Oils  " 
(p.  177).  The  alcohol  and  benzene  having  been 
removed  by  evaporation,  the  residual  soap  is  dissolved 
in  water,  and  the  solution  separated  from  unsaponifi- 
able  matter  by  extraction  with  ether.  The  soap 
solution  is  then  evaporated  to  dryness,  redissolved 
in  water,  and  the  solution  acidified  with  hydrochloric 
acid ;  the  precipitated  acids  are  extracted  with  ether, 
after  which  the  aqueous  layer  is  separated  off, 
neutralised  and  evaporated  to  about  25  c.c.  On 
acidifying  again  and  extracting  with  ether,  the  rest  of 
the  fatty  acids  will  be  separated.  The  ethereal  extracts 
are  united  and  evaporated ;  the  residue  may  then  be 
tested  for  rosin  acids  by  the  Liebermann-Storch  reaction. 
If  desired,  the  rosin  acids  may  be  estimated  by  Twitchell's 
process,  as  described  in  Chapter  IV.,  p.  144,  in  the 
mixture  of  fatty  and  rosin  acids  obtained,  as  just  de- 
scribed, from  a  given  weight  of  the  oil  (sufficient  to  give 
about  5  grams  of  mixed  acids).  It  should  be  noted  that 
rosin  acids  which  have  been  treated  with  hydrochloric 
acid,  as  in  the  Twitchell  process,  no  longer  give  the 
Liebermann-Storch  reaction. 

Rosin  Oil. — Rosin  oil  is  obtained  from  rosin,  or  colo- 
phony, by  destructive  distillation.  It  always  contains 
varying  quantities,  up  to  about  30  per  cent.,  of  unchanged 
rosin  acids  which  have  been  carried  over  mechanically 
in  the  distillation,  and  may,  therefore,  be  tested  for  in 
lubricating  oils,  as  described  above,  by  the  Liebermann- 
Storch  reaction.  Rosin  oil  gives  a  violet  coloration  when 
shaken  with  a  drop  of  stannic  bromide  (Allen). 

There  is  no  thoroughly  satisfactory  method  for  the 
quantitative  estimation  of  rosin  oils  in  presence  of 
mineral  oils,  the  best  methods  available  being  those  of 

12 2 


i8o      Industrial  Organic  Analysis 

Valenta  and  Storch,  which  are  based  on  the  solubility  of 
rosin  oil  in  glacial  acetic  acid  and  96  per  cent,  alcohol, 
respectively.  These,  and  also  other  methods  for  the 
detection  of  rosin  oils,  will  be  found  described  in  the 
works  mentioned  at  the  end  of  this  chapter. 

Rosin  oils  are  generally  used  in  axle  greases,  together 
with  soap  and  water,  or  as  insulating  material  for 
electrical  machines.  Owing  to  their  tendency  to  resinify 
when  exposed  to  the  air  in  thin  layers  at  elevated 
temperatures,  they  are  unsuitable  for  use  in  the  better 
class  lubricating  oils.  Rosin  also  has  a  tendency  to 
cause  gumming,  and,  like  rosin  oil,  must  be  regarded  as 
an  adulterant  if  found  to  be  present  in  better  class 
lubricating  oils. 

Lubricating  oils  are  sometimes  tested  for  their  gum- 
ming tendency  by  spreading  them  in  thin  layers  on  glass 
plates  and  keeping  them  at  50°  or  100°  C.,  the  film  being 
examined  on  cooling,  from  day  to  day.  If  rosin  oil  be 
compared  in  this  way  with  a  mineral  lubricating  oil  of 
good  quality,  the  superiority  of  the  latter  as  regards 
gumming  tendency  may  be  shown. 

Tar  Oils. — The  tar  oils  which  are  used  for  the  adultera- 
tion of  mineral  lubricating  oils  are  usually  the  heavy 
dark  anthracene  mother  liquor  oils  (see  Chapter  II.),  of 
specific  gravity  over  i-oo.  If  present  in  fairly  large 
quantity,  therefore,  they  may  sometimes  be  recognised 
by  the  somewhat  high  specific  gravity  of  the  sample. 
According  to  Holde,  the  following  properties  of  tar  oils 
enable  them  to  be  recognised  in  presence  of  mineral  oils. 
They  are  completely  soluble  in  alcohol  at  the  ordinary 
temperature,  giving  dark-coloured  solutions,  and  their 
smell  is  generally  creosote  like.  When  heated  on  the 
water  bath  with  concentrated  sulphuric  acid,  they  react 
to  form  water  soluble  products.  With  nitric  acid  of 
specific  gravity  1-45,  they  react,  vigorously,  often  with  ex- 
plosive violence,  with  the  formation  of  nitro  compounds. 


Petroleum  and  its  Products      181 

Lignite  Oils. — The  lignite  oils  of  high  boiling  point  and 
specific  gravity  0-9  to  0-97,  which  may  be  used  for  the 
adulteration  of  mineral  oils,  resemble  the  coal  tar  oils  in 
having  a  creosote  like  smell ;  they  are  soluble  in  twice 
their  volume  of  alcohol  at  the  ordinary  temperature,  to 
the  extent  of  22  to  62  per  cent.  They  contain  unusually 
large  amounts  of  sulphur  compounds,  and  react  with 
concentrated  nitric  acid  less  vigorously  than  the  coal  tar 
oils,  but  more  vigorously  than  the  mineral  oils.  They  are 
not  so  easily  recognised  in  mixtures  as  the  coal  tar  oils. 

The  detection  of  other  possible  constituents  of  lubrica- 
ting oils,  such  as  dissolved  rubber,  deblooming  agents, 
etc.,  will  be  found  described  in  the  works  mentioned 
below. 

REFERENCES. 

"  Untersuchung  der  Mineralole  und  Fette,"  by  D.  Holde. 
1909,  J.  Springer,  Berlin. 

"  Das  Erdol  und  Seine  Verwandten,"  by  H.  von  Hofer.  1912, 
Fr.  Vieweg,  Brunswick. 

"  Die  Schmiermittel,  Methoden  zu  Ihre  Untersuchung  und 
Wertbestimmung,"  by  Grossmann.  1909,  Kriedel's  Verlag, 
Wiesbaden. 

"  The  Chemical  Technology  and  Analysis  of  Oils,  Fats  and 
Waxes,"  by  J.  Lewkowitsch,  3  vols.  1910,  Macmillan  &  Co., 
London. 

"  Engineering  Chemistry,"  by  Stillmann.  1910,  Williams  and 
Norgate,  London,  and  the  Chemical  Publishing  Co.,  Fasten,  Pa. 

"  Das  Erdol,"  by  Engler  and  Hofer,  3  vols.  1908  to  1911, 
S.  Hirzel,  Leipzig. 

"  Laboratoriumsbusch  fur  der  Erdolindustrie,"  by  Kissling. 
1908,  Wilhelm  Knapp,  Halle. 

"  Handbook  on  Petroleum,"  by  J.  H.  Thomson  and  Boverton 
Redwood.  1906,  C.  Griffin  &  Co.,  London. 

Allen's  "  Commercial  Organic  Analysis  "  (revised  edition), 
Vol.  III.  1910,  J.  and  A.  Churchill,  London. 

"  Chemisch-Technische  Analyse,"  Vol.  I.,  by  Post.  1908, 
Fr.  Vieweg,  Brunswick. 

"  Chemisch-Technische  Untersuchungsmethoden,"  Vol.  III., 
by  Lunge.  1911,  J.  Springer,  Berlin. 


CHAPTER   VI 
MILK  AND  BUTTER 

INTRODUCTORY 

ON  account  of  the  enormous  value  of  the  milk  of  the 
cow  as  an  article  of  food,  both  in  its  original  state  and 
when  manufactured  into  such  products  as  butter  or 
cheese,  its  complex  nature  and  its  susceptibility  to 
decomposition  through  the  agency  of  the  manifold 
species  of  micro-organisms  for  which  it  forms  an  excellent 
nutrient  medium,  the  examination  of  this  product  may 
well  be  said  to  form  one  of  the  most  important  and 
interesting  chapters  in  the  chemistry  of  foods.  The 
examination  of  whole  milk,  skim  milk,  cream,  condensed 
milk  and  butter  will  be  described  below. 

Ordinary  cow's  milk  of  commerce  is  a  white  opaque 
emulsion  of  fat  in  water  which  contains  lactose,  saline 
matter  and  other  substances  in  solution,  and  casein  in  a 
state  of  colloidal  suspension.  Its  bluish  or  yellowish 
tinge  depends  on  the  amount  of  fat  present,  or  on  the 
presence  of  colouring  matter,  either  natural  or  artificial, 
in  the  latter. 

Fresh  milk  has  an  amphoteric  reaction,  turning  blue 
litmus  slightly  red,  and  red  litmus  slightly  blue  ;  this  is 
owing  to  the  presence  of  acid  phosphates  ;  on  an  average, 
100  c.c.  of  fresh  milk  will  be  found  to  require  about  30  c.c. 
of  decinormal  sodium  hydroxide  solution  for  neutralisa- 
tion in  presence  of  phenol  phthalein,  and  40  c.c.  of  deci- 
normal sulphuric  acid  in  presence  of  litmus  as  indicator. 
On  standing,  milk  which  has  not  previously  been  heated 
almost  invariably  becomes  decidedly  acid  owing  to  the 
conversion  of  part  of  the  lactose  into  lactic  acid  by  certain 
micro-organisms. 

The  most  important  constituents  of  milk  are  water, 


Milk  and  Butter  183 

fat,  proteins,  lactose  and  inorganic  salts.  Roughly, 
nine-tenths  of  the  weight  is  water  and  one-tenth  solid 
matter.  The  following  data,  showing  the  general  varia- 
tions in  the  composition  of  cow's  milk,  have  been  com- 
piled by  Barthel  from  many  different  sources  :— 

Specific  gravity           . .  . .  1-029  t°  1*034 

Fat        . .         . .         . .  . .  2-5  ,,  4-5      per  cent. 

Total  solids      . .          . .  . .  10-3  ,,  14-7          ,, 

Solids  less  fat              . .  . .  7-8  „  10-2          ,, 

Fat  in  solids     . .          . .  . .  19-9  ,,  32-0          ,, 

Specific  gravity  of  solids  . .  1-3  ,,  1-4            „ 

Fleischmann  gives  the  following  figures,  which  are 
based  on  numerous  analyses  : — 

Water  . .         . .  86-5  to  89-5  per  cent. 

Fat      ..         ..         ..         2-7  „     4-3 

Proteins          . .          . .         3-0  „     4-0        „ 

Lactose  . .         . .         3-6  „     5-5 

Ash 0-6  „     0-9        „ 

All  the  above  values  are  based  on  analyses  of  samples 
of  mixed  milk  from  herds  of  cows  ;  if  based  on  analyses 
of  samples  from  individual  animals,  the  variations  would 
have  been  far  greater.  The  composition  of  milk  varies, 
not  only  with  the  age,  state  of  health,  and  breed  of  the 
animal,  but  also  with  the  district,  climatic  conditions, 
time  of  the  year,  general  treatment,  method  of  feeding 
and  other  factors.  It  therefore  follows  that  the  stand- 
ard values  used  for  purposes  of  comparison  should  be 
based  on  as  large  a  number  of  analyses  as  possible,  while 
the  values  adopted  as  the  minimum  permissible  amounts 
of  fat  and  solids  less  fat  may  vary  in  different  countries. 

Before  proceeding  to  describe  analytical  methods,  a 
short  account  will  be  given  of  the  principal  constituents 
of  milk  and  butter. 

THE  CONSTITUENTS  OF  MILK. 

Fat. — As  stated  above,  this  constituent  is  present  in 
the  form  of  an  emulsion  ;  the  globules  of  fat  generally 
measure  from  0-0016  to  o-oi  mm.  in  diameter,  and 


184      Industrial  Organic  Analysis 

number,  on  an  average,  2\  to  3  millions  per  cubic  milli- 
metre. In  milk,  the  fat  is  present  in  the  liquid,  super- 
cooled state  ;  on  transformation  into  butter  fat,  it  takes 
up  the  solid  form  which  melts  from  31°  to  36°.  This 
change  is  effected  by  thoroughly  agitating  cream  which 
contains  about  25  to  35  per  cent,  of  fat,  and  which  has 
usually  previously  been  soured  or  "  ripened  "  by  lactic 
acid  producing  organisms,  in  specially  constructed 
churns,  at  temperatures  of  about  10°  to  15°  ;  the  solid 
fat  coalesces  and  separates  from  the  buttermilk  in  the 
form  of  small  pellets,  which,  after  washing  with  water, 
are  worked  by  a  process  of  kneading  into  butter  contain- 
ing some  12  to  16  per  cent,  of  water.  The  composition 
of  butter  will  be  further  treated  of  below. 

When  milk  is  allowed  to  stand  at  rest,  the  fat  rises  to 
the  top,  forming  a  layer  of  milk  rich  in  fat,  known  as 
cream.  In  modern  dairy  practice,  milk  is  separated  in 
centrifugal  machines  from  which  it  is  possible  to  obtain 
cream  containing  anything  up  to  80  per  cent,  of  fat  at 
will ; .  the  skim  milk  obtained  contains  a  few  tenths  per 
cent.,  or  even  less,  of  fat. 

Milk  fat  is  peculiar  in  containing  an  unusually  large 
number  of  different  glycerides  which  include  notable 
proportions  of  the  glycerides  of  the  lower  fatty  acids, 
especially  butyric  acid.  As  is  pointed  out  in  Chapter  III., 
the  latter  circumstance  is  taken  advantage  of  in  the 
estimation  of  butter  fat  by  the  Reichert-Wollny  and 
other  processes  in  which  the  volatile  fatty  acids  obtained 
from  a  given  weight  of  the  fat  are  estimated.  Besides 
the  lower  fatty  acids,  milk  or  butter  fat  yields  on  hydro- 
lysis higher  fatty  acids  such  as  palmitic,  stearic  and 
oleic  acids  (see  Chapter  III.,  introductory),  the  acids 
of  intermediate  molecular  weight  being  obtained  in 
relatively  small  proportion.  Milk  fat  is  the  most  valu- 
able fat  known,  and  is,  economically,  the  most  important 
constituent  of  milk.  The  commercial  valuation  of  milk 
is  based  very  largely  on  its  fat  content,  while  from  the  fat 
percentage  of  the  milk  yielded  by  individual  cows  it  is 
possible  to  calculate  the  amount  of  butter  obtainable  in 
relation  to  the  fodder  consumed  by  the  animals. 

Although  the  fat  in  milk  from  individual  cows  may 


Milk  and  Butter  185 

vary  from  i  to  8  per  cent,  when  exceptional  cases  are 
included,  the  variations  for  mixed  milk  from  herds  of 
cows,  taken  all  the  year  round,  usually  fall  between 
2-5  and  4-5  per  cent.  In  order  to  illustrate  the  influence 
of  the  breed  of  the  animals,  it  may  be  mentioned  that  the 
milk  from  Jersey  cows  averages  5  per  cent,  of  fat, 
while  that,  from  Dutch  cows  only  averages  3-.!  to 
3-2  per  cent.  The  standards  adopted  by  different 
countries  in  fixing  the  minimum  permissible  proportion  of 
fat  will  naturally  vary  to  some  extent.  In  the  United 
Kingdom  it  is  generally  presumed  that  genuine  milk 
should  contain  at  least  3  per  cent,  of  fat,  and  accordingly 
the  Sale  of  Milk  Regulations  of  1901  provide  that  milk 
containing  less  than  3  per  cent,  of  milk  fat  (or  less  than 
8-5  per  cent,  of  milk  solids  other  than  fat)  is  to  be  pre- 
sumed not  to  be  genuine,  unless  the  contrary  be  proved. 
Thus,  no  absolute  standard  is  fixed,  but  the  burden  of 
proof  would  lie  with  the  defendant  in  the  case  of  a 
prosecution. 

Proteins.— The  total  proteins  of  milk  amount,  on  an 
average,  to  about  3-5  per  cent.,  of  which  about  2-9  per 
cent,  is  casein,  and  0-6  per  cent,  milk  albumin  together 
with  traces  of  milk  globulin.  Casein  belongs  to  the 
phosphoproteins,  but  differs  from  other  phosphorus- 
containing  proteins  in  that  it  yields  no  xanthine  and 
pyrimidine  bases,  or  pentoses,  on  hydrolysis.  Owing  to 
the  absence  of  carbohydrate  groupings,  it  fails  to  give 
the  Molisch  reaction  ;  this  test  (which  may  be  carried 
out  with  egg  albumen)  consists  in  adding  to  an  aqueous 
solution  of  the  protein  a  few  drops  of  an  alcoholic  solu- 
tion of  a  naphthol  and  then  concentrated  sulphuric  acid  ; 
if  carbohydrate  groupings  are  present,  as  is  the  case  with 
egg  albumen,  a  violet  coloration  will  be  formed  at  the 
surface  of  contact  of  the  acid  and  aqueous  layers. 

On  hydrolysis  with  baryta  solution,  casein  yields, 
besides  carbon  dioxide  and  ammonia,  a  number  of  amino 
acids  which  consist,  in  the  main,  of  monoamino  fatty 
acids  and  cyclic  derivatives  of  the  latter.  Pure  casein  is 
a  white  powder,  insoluble  in  water  and  soluble  in  dilute 
alkali  or  acid  solutions.  It  thus  possesses  feebly  basic 
and  acidic  properties.  In  milk,  casein  exists  in  the  form 


1 86      Industrial  Organic  Analysis 

of  a  calcium  salt,  which,  being  present  in  the  colloidal 
state,  can  only  be  filtered  off  by  passing  the  milk  through 
a  filter  of  unglazed  porcelain  ;  in  this  way  the  total 
proteins  of  the  milk  may  be  obtained  as  a  white  mass. 

When  milk  is  treated  with  a  small  proportion  of  acetic 
or  mineral  acid,  or  allowed  to  become  sour  through  the 
agency  of  lactic  acid  producing  organisms,  it  thickens 
owing  to  the  precipitation  of  the  casein  which  has  been 
liberated  from  its  calcium  derivative.  On  warming  the 
acidified  solution,  the  casein  coagulates,  leaving  a  slightly 
turbid,  yellowish  whey,  while  on  addition  of  excess  of 
acid  it  is  redissolved.  On  the  addition  of  rennet, 
tannic  acid,  alcohol  or  inorganic  salts,  such  as  sodium 
chloride,  copper  sulphate,  alum,  etc.,  to  milk,  the  casein 
is  precipitated  unchanged,  in  the  form  of  the  calcium 
derivative.  In  the  latter  respect,  it  resembles  the 
majority  of  proteins,  which  may  usually  be  separated 
and  purified  by  the  salting  out  of  their  aqueous  solutions 
at  various  definite  concentrations  and  temperatures. 

Milk  albumin  is  very  similar  in  composition  and  pro- 
perties to  the  albumin  of  the  blood.  It  is  not  precipi- 
tated with  the  casein  on  the  acidification  of  milk,  and 
further  differs  from  casein  in  not  being  salted  out  from 
neutral  solution  on  saturation  with  magnesium  or 
ammonium  sulphate.  It  is,  however,  precipitated  from 
slightly  acid  solution  on  saturation  with  the  above  salts. 

Milk  Sugar. — Milk  sugar,  or  lactose,  is  a  disaccharide 
which,  on  hydrolysis  with  dilute  mineral  acid,  yields  the 
hexoses  galactose  and  glucose  in  equal  amounts.  It 
reduces  Fehling's  solution,  and  is  dextrorotatory  in 
aqueous  solution,  the  hydrated  lactose  C^H^On,  H2O 
having  a  specific  rotation  of  [a]D+  52-53°  at  20°  C.  As 
will  be  shown  below,  the  two  last-mentioned  properties 
are  made  use  of  in  the  quantitative  estimation  of  lactose. 

The  fermentation  changes  undergone  by  lactose  play 
an  important  part  both  in  the  preservation  of  milk  and 
the  manufacture  of  butter  from  cream.  Lactose  is  not 
fermented  by  the  common  yeasts,  most  of  which  attack 
cane  sugar  ;  by  far  the  most  important  fermentation 
change  which  it  undergoes  is  its  transformation  into 
lactic  acid  (C12H2Ai  +  H2O  =  4  C3H6O3). 


Milk  and  Butter  187 

This  change  is  brought  about  through  the  agency  of 
different  species  of  micro-organisms  which  occur  in  milk, 
and  is,  under  normal  conditions,  the  first  change  in  the 
decomposition   of  unheated  milk   by  micro-organisms. 
As  was  pointed  out  above,  the  milk  becomes  coagulated 
owing  to  the  precipitation  of  the  casein  by  the  acid.     In 
modern  dairy  practice,  cream  is  soured  for  butter-making 
by  first  pasteurising  it,  i.e.,  heating  to  about  70°  to  85°, 
in  order  to  destroy  most  of  the  micro-organisms  present, 
and  then,   after  cooling  to  the  ordinary  temperature, 
infected  with  a  pure  culture  of  a  lactic  acid  producing 
organism,  such  as  Streptococcus  Lacticm.     By  this  means 
it   is   possible   to  obtain   more   uniform  results  and   a 
better   flavour   in   the   butter  than   if,  as   in   the  old 
process,    the   cream  were   allowed   to   ripen   spontane- 
ously through  the  agency  of   such  naturally  occurring 
bacteria  as  may  obtain  a  predominating  influence.     The 
formation   of  lactic   acid  is,   quantitatively,   the   chief 
change  which  occurs  in  the  souring  of  milk  or  cream  ;  at 
the  same  time,  however,  traces  of  aroma-producing  sub- 
stances are  formed  which  play  an  important  part  in  the 
flavouring  of  the  butter  fat. 

Mineral  Matter.  —  The  analysis  of  the  ash  of  milk, 
allowing  for  the  phosphorus  contained  in  the  proteins, 
and  the  carbonate  formed  from  the  organic  matter  on 
incineration,  shows  the  mineral  matter  of  milk  to  be 
composed  as  follows  (Soldner,  quoted  from  Barthel)  :— 

Per  cent. 
Sodium  chloride          .  .          .  .          .  .         10-62 

Potassium  chloride     .  .          .  .          .  .  9-16 

Mono  potassium  phosphate  .  .          .  .         1277 

Dipotassium  phosphate          .  .          .  .  9-22 

Potassium  citrate        .  .          .  .          .  .  5-47 

Dimagnesium  phosphate        .  .          .  .  3-71 

Magnesium  citrate      .  .          .  .          .  .  4*05 

Dicalcium  phosphate  .  .          .  .          .  .  7-42 

Tricalcium  phosphate  .  .          .  .  8-90 

Calcium  citrate  .  .          .  .          .  .         23-55 

Calcium  combined  with  casein          .  .  5-13 


Total..  100-00 


1 88      Industrial  Organic  Analysis 

All  the  above  constituents  exist  in  the  dissolved  state, 
with  the  exception  of  the  calcium  combined  with  the 
casein,  which  exists  in  the  colloidal  state,  and  part  of  the 
calcium  phosphate,  which,  although  existing  in  the  solid 
form,  is  exceedingly  finely  divided. 

Other  Constituents. — Lecithin,  a  glycero-phosphate  of 
the  trimethyl-ammonium  base,  choline,  occurs,  on  an 
average,  to  the  extent  of  about  0-065  per  cent,  in  milk. 
The  presence  of  this  substance  is  said  to  cause  the 
"  browning  "  which  takes  place  on  heating  butter. 

Enzymes  of  different  kinds  are  contained  in  milk  ; 
these  include  oxidases,  reductases,  catalases,  and  protein 
and  fat  hydrolysing  enzymes.  The  significance  of  some 
of  these  enzymes,  some  of  which  are  probably  produced 
by  the  bacteria  in  the  milk,  will  be  dealt  with  later. 

Dissolved  gases,  consisting  of  oxygen,  carbon  dioxide, 
etc. 

Among  foreign  bodies,  constantly  occurring  in  milk, 
are  white  blood  corpuscles  or  leucocytes,  bacteria, 
yeasts,  moulds  and  their  spores  and  dirt. 

The  number  of  micro-organisms  and  amount  of  dirt 
contained  in  new  milk  depend  on  the  cleanliness  of  the 
cows,  the  stable,  and  the  vessels  in  which  the  milk  is 
collected,  as  well  as  on  the  care  which  is  taken  in  the 
subsequent  treatment  of  the  milk.  Thus,  while  milk 
from  clean  cows,  collected  in  steam-sterilised  vessels,  may 
contain  only  a  few  hundred  bacteria  per  cubic  centi- 
metre immediately  after  milking,  milk  from  dirty  cows, 
collected  in  dirty  vesssels,  may  contain  several  hundred 
thousand  bacteria  per  cubic  centimetre.  The  dirt,  which 
may  consist  of  hairs  of  animals  and  human  beings,  earth, 
fodder,  fibres,  sand,  parts  of  insects,  etc.,  may  amount  to 
as  much  as  0-03  to  0-25  per  cent,  of  the  milk,  but  should, 
in  general,  be  below  o-oi  per  cent. 

THE  CONSTITUENTS  OF  BUTTER. 

As  will  be  gathered  from  the  above  remarks  on  the 
production  of  butter,  the  latter  contains  the  constituents 
of  milk  in  altered  proportions.  The  percentage  of  fat  in 
butter  will  generally  lie  between  83  and  85  per  cent., 


Milk  and  Butter  189 

water  between  12  and  15  per  cent.,  while  the  proteins 
generally  amount  to  about  0-6  to  0-9  per  cent.,  sugar  and 
lactic  acid  to  0*4  per  cent.,  and  ash,  i.e.,  inorganic  salts, 
to  o-i  or,o-2  per  cent. 

The  fat,  which  differs  from  milk  fat  in  being  in  the 
solid  state,  is  present  in  the  form  of  minute  globules, 
emulsified  in  the  aqueous  serum  which  holds  the  proteins, 
or  curd,  in  suspension  and  the  other  constituents,  such 
as  inorganic  salts  and  sugar,  in  solution.  The  content  of 
curd  may  vary  somewhat  according  to  the  method  of 
manufacture  of  the  butter ;  thus,  Storch  gives  the 
average  protein  content  of  butter  from  fresh  cream  as 
0-64,  and  of  butter  from  ripened  or  soured  cream  as  0-84 
per  cent.  A  certain  amount  of  salt  is  almost  invariably 
added  to  butter  as  a  flavouring  medium  and  as  a  pre- 
servative, in  order  to  check  the  development  of  micro- 
organisms which  may  turn  it  rancid.  The  amount  of  salt 
added  may  vary  from  I  to  5  per  cent.,  the  so-called  fresh 
or  mild  cured  butters  containing  less  than  ij  per  cent. 
Brine-pickled  butters  may  contain  as  much  as  6  to  9  per 
cent,  of  salt.  Boric  acid  or  borax  preservative  may  be 
added  in  amounts  up  to  one  half  per  cent.,  expressed  as 
boric  acid. 

THE  PHYSICAL  AND  CHEMICAL  EXAMINATION  OF  MILK, 
CREAM  AND  SKIM  MILK. 

Sampling. — For  a  complete  examination,  exclusive  of 
the  estimation  of  dirt,  about  500  c.c.  will  be  required. 
Owing  to  the  tendency  of  the  fat  to  collect  at  the  top,  the 
sample  should  always  be  taken  after  thoroughly  mixing 
the  bulk  of  the  milk.  The  sample  bottle  and  cork  should 
be  perfectly  clean,  and  sterilised  by  wrapping  in  filter 
paper  and  heating  in  an  air  oven  to  about  130°  to  150° 
for  half  an  hour.  The  bottle  should  not  be  filled  more 
than  about  three-quarters  full,  in  order  that  the  cream 
may  be  mixed  into  the  bulk  of  the  milk  by  shaking, 
immediately  before  examination. 


190      Industrial  Organic  Analysis 

If  the  analysis  cannot  be  proceeded  with  at  once,  the 
sample  is  best  preserved  by  keeping  it  in  a  cool  place, 
preferably  in  an  ice  chest.  The  addition  of  0-5  gram  of 
potassium  dichromate,  or  i  c.c.  of  a  40  per  cent,  solution 
of  formaldehyde  per  litre,  will  preserve  milk  for  a  pro- 
longed period  ;  it  is,  however,  advisable  to  avoid,  as  far 
as  possible,  tampering  with  the  sample  in  any  way  ; 
further,  the  addition  of  dichromate  will  naturally  affect 
the  specific  gravity  and  content  of  ash  and  solids  less  fat, 
while  the  addition  of  formaldehyde- may  interfere  with 
the  determination  of  fat  by  centrifugal  methods. 

Immediately  before  the  analysis  the  sample  bottle 
should  be  repeatedly  turned  in  order  to  mix  the  cream 
with  the  rest  of  the  milk,  heating,  if  necessary,  to  about 
40°.  The  milk  is  then  poured  into  a  clean  beaker  ;  if 
any  cream  remains  in  the  bottle,  the  milk  should  be 
returned  to  it  and  shaken  until  only  a  trace  of  cream 
adheres  to  the  sides.  Every  time  a  portion  of  the  sample 
is  to  be  weighed  or  measured  off,  it  should  be  mixed  by 
pouring  from  one  vessel  to  another.  In  case  the  sample 
for  analysis  should  be  soured  and  coagulated,  a  special 
method  of  treatment  is  recommended  by  Weibull,  for 
which  see  Barthel's  "  Analysis  of  Milk  and  Dairy  Pro- 
ducts." 

PHYSICAL  EXAMINATION. 

The  analytical  processes  described  under  this  heading 
include  the  determination  of  specific  gravity  and  dirt  or 
foreign  matter.  In  appearance,  the  milk  should  be  fully 
opaque  and  homogeneous,  and  on  standing  at  rest,  it 
should  form  a  well-defined  layer  of  cream  ;  a  collection 
of  flaky  particles  indicates  either  udder  disease  or  that 
the  milk  is  so  old  that  bacterial  decomposition  has  set  in. 
Skim  milk  will,  of  course,  separate  no  cream  layer.  A 
sour  taste  or  smell  indicates  that  lactic  acid-producing 
organisms  have  become  active,  while  a  bitter  or  saltish 


Milk  and  Butter 


191 


taste  indicates  either  that  the  milk  is  derived  from  cows 
which  have  been  improperly  fed  or  are  suffering  from 
udder  disease,  or  else  that  the  proteins  of  the  milk  are 
being  decomposed  by  bacteria  with  the  formation  of 
peptones.  A  metallic  taste  is  generally  due  to  the  use 
of  untinned  vessels,  while  the  use  of 
certain  fodders,  such  as  mangel-wurzels 
or  turnips,  also  produces  characteristic 
tastes  in  the  milk.  Milk  very  easily 
acquires  the  taste  and  smell  of  materials 
such  as  carbolic  acid  or  dung  if  left  in 
their  neighbourhood  for  some  time. 

Specific  Gravity. — The  specific  gravity 
of  milk  is  determined  with  the  object  of 
detecting  adulteration,  either  by  addi- 
tion of  water  or  removal  of  cream  ;  the 
interpretation  of  the  results  will  be  left 
to  a  subsequent  section.  (See  "Solids 
less  Fat.") 

The  determination  should  not  be 
made  within  a  period  of  3  hours  of  the 
milking,  owing  to  the  slight  increase  in 
specific  gravity  which  takes  place  at 
first,  probably  owing  to  some  change  in 
the  colloidal  state  of  the  casein.  The 
determination  may  be  carried  out  in 
the  usual  way  by  means  of  the  West- 
phal  balance,  specific  gravity  bottle,  or, 
as  is  more  commonly  the  case,  with  a 

float  specially  designed  for  the  purpose,  FIG.  n. Soxhlet's 

known  as  a  lactometer.  Soxhlet's  lacto-  Lactometer, 
meter  is  furnished  with  a  thermometer,  and  is  graduated 
to  show  specific  gravities  from  1-024  to  1-038.  The 
length  of  the  scale  divisions  is  so  adjusted  that  the 
fourth  decimal  place  may  be  estimated.  The  instru- 
ment is  adjusted  to  15°  C.,  the  specific  gravity  found 
indicating  the  ratio  of  a  given  weight  of  the  milk  to 


192      Industrial  Organic  Analysis 

that  of  an  equal  weight  of  water  at  this  temperature. 
Readings  may  be  reduced  to  15°  by  adding  two  units  to 
the  fourth  decimal  place  for  each  degree  above  15°,  and 
subtracting  two  units  for  each  degree  below ;  these 
corrections  hold  good  for  temperatures  between  10°  and 
20°  C. 

Before  taking  the  specific  gravity  the  sample  should 
be  well  mixed  and  then  poured  into  a  cylinder  of  suitable 
size,  avoiding  the  formation  of  froth  on  the  surface. 
The  float  is  carefully  lowered  into  the  milk  until  the 
division  1-030  coincides  with  the  surface,  and  then  allowed 
to  come  to  equilibrium  of  its  own  accord.  Keeping  the 
eye  on  a  level  with  the  surface  of  the  milk,  the  reading 
is  taken  at  the  highest  point  at  which  the  milk  is  seen  to 
rise  up  the  spindle.  The  specific  gravity  of  whole  milk 
usually  lies  between  1-029  and  1*034,  or>  as  *s  sometimes 
stated,  between  29°  and  34°.  Skim  milk  has  a  somewhat 
higher  specific  gravity,  usually  between  1-035  and 
1*037,  owing  to  the  absence  of  fat. 

Dirt  and  Foreign  Matter. — The  following  method  by 
which  the  foreign  insoluble  matter  is  estimated  in  milk 
is  due  to  Stutzer  :  the  apparatus  required  consists  of  a 
bottle  of  about  a  litre  capacity,  having  a  tapering  neck, 
connected  by  means  of  a  wide  piece  of  rubber  tube  to  a 
strong  test  tube  without  a  rim.  For  the  purpose  an 
ordinary  Rhenish  wine  bottle  may  be  used.  The  bottle  is 
charged  with  a  litre  of  milk,  connected  with  the  test 
tube,  and  the  whole  is  allowed  to  stand  for  2  hours 
in  an  inverted  position,  so  that  any  sediment  will  collect 
in  the  test  tube.  The  rubber  connection  is  then  pinched 
together  with  a  clip,  and  the  bottle  is  disconnected  ;  the 
milk  is  decanted  from  the  'sediment  in  the  test  tube, 
which  is  washed  several  times  with  distilled  water 
acidified  with  a  little  hydrochloric  acid,  in  order 


Milk  and  Butter  193 

to  dissolve  the  calcium  phosphate,  which  is  a  normal 
constituent  of  the  milk.  The  sediment  is  collected  on  a 
Gooch  crucible,  washed  with  water  until  the  nitrate  is  no 
longer  opalescent,  then  with  alcohol  and  ether,  dried 
at  1 00°  and  weighed. 

Berch  recommends  that  the  milk  should  be  preserved 
with  formalin  (see  under  "  Sampling  "),  in  order  that  it 
may  be  allowed  to  settle  for  24  hours.  Gerber  has 
devised  a  modification  of  the  above  method  by  which  a 
number  of  determinations  may  be  carried  out  simul- 
taneously. (See  Barthel's  "  Analysis  of  Milk  and  Dairy 
Products.") 

Good  clean  milk  should  contain  only  from  5  to  10  milli- 
grams of  dirt  per  litre,  though  milk  of  commerce  often 
contains  considerably  more.  The  figures  obtained  by 
this  method  should  not  be  taken  too  literally,  as  some  of 
the  original  dirt  may  have  dissolved  in  the  milk,  some 
adheres  to  the  fat  globules  and  is  carried  to  the  surface, 
while  a  further  quantity  may  dissolve  in  the  alcohol  and 
ether.  Generally  speaking,  however,  the  method  gives 
a  fair  idea  of  the  care  which  has  been  taken  in  the  treat- 
ment of  the  milk. 

A  method  by  which  the  relative  number  of  micro- 
organisms may  be  roughly  estimated  with  a  view  to 
forming  an  opinion  as  to  the  care  and  cleanliness  in  the 
previous  treatment  of  the  milk  is  described  later. 

CHEMICAL  EXAMINATION. 

Under  this  heading  will  be  described  the  determination 
of  the  acidity  of  milk,  fat  in  whole  milk,  skim  milk  and 
cream,  proteins,  including  casein  and  albumin,  lactose 
and  ash  in  milk,  after  which  the  analytical  results  will 
be  discussed  from  the  point  of  view  of  the  detection  of 
the  adulteration  of  milk.  Finally,  tests  for  the  pasteur- 
ising and  bacterial  contents  of  milk  will  be  dealt  with. 
The  examination  of  milk,  as  well  as  butter,  for  preserva- 
tives will  be  considered  in  Chapter  VIII. 

I.O.A.  13 


194      Industrial  Organic  Analysis 

Acidity. — As  mentioned  above,  the  production  of 
lactic  acid  from  the  lactose  of  the  milk  is  usually  the  first 
change  produced  through  the  agency  of  the  micro- 
organisms which  normally  occur  in  milk.  A  determina- 
tion of  the  acidity  of  the  sample  will,  therefore,  enable 
the  analyst  to  form  an  opinion  as  to  whether  it  is  fresh 
or  not.  The  acidity  of  cream  is  also  determined  as 
described  below,  as  a  check  on  the  souring  process 
in  butter  making.  As  stated  above,  normal  fresh  milk 
has  an  amphoteric  reaction.  A  decidedly  acid  reaction 
towards  litmus  will  show  that  souring  has  already  com- 
menced, while  a  decidedly  alkaline  reaction  will  indicate 
that  the  milk  is  derived  from  sick  cows,  or  that  it  has 
been  diluted  with  water.  Empirical  tests  for  the  fresh- 
ness of  milk  are  the  alcohol  and  the  boiling  tests,  which 
depend  on  the  precipitation  of  the  casein,  i.e.,  the  curd- 
ling of  the  milk  which  takes  place  on  adding  alcohol  or 
boiling,  when  the  milk  contains  a  certain  proportion  of 
free  acid. 

The  alcohol  test  is  carried  out  by  adding  to  10  c.c.  of 
milk  an  equal  volume  of  68  per  cent,  alcohol  and  shaking 
well ;  fresh  milk  shows  no  change,  while  milk  which  has 
8-5  or  more  Soxhlet-Henkel  degrees  of  acidity  (see 
below),  or  milk  from  animals  with  diseased  udders,  shows 
a  more  or  less  distinct  coagulation.  Milk  for  infants 
should  not  coagulate  on  addition  of  twice  its  volume  of 
68  per  cent,  alcohol. 

Coagulation  on  warming  in  a  test  tube  shows  that  the 
milk  contains  upwards  of  0-26  per  cent,  of  lactic  acid. 

The  acidity  of  milk  is  determined,  according  to 
Soxhlet  and  Henkel,  by  adding  to  50  c.c.  of  milk,  2  c.c.  of 
a  2  per  cent,  solution  of  phenol  phthalein  and  titrating 
with  quarter-normal  sodium  hydroxide  solution  till  a 
faint  pink  colour  remains  on  shaking.  The  number  of 
cubic  centimetres  required  in  the  titration  gives  the 
acidity  of  "the  milk  in  Soxhlet-Henkel  degrees.  It  is,  of 
course,  not  necessary  to  adhere  to  the  above  proportions ; 


Milk  and  Butter  195 

thus  it  may  be  found  convenient  to  use  a  smaller  quantity 
of  milk,  and  weaker  standard  alkali  solution  for  the 
titration. 

Fresh  cow's  milk  shows,  on  an  average,  3-5°  of  acidity 
according  to  the  above  method.  Milk  in  which  the  pro- 
cess of  spontaneous  souring  has  been  allowed  to  com- 
plete itself  will  show  an  acidity  of  over  15°,  while  soured 
cream  for  butter  making  should  have  an  acidity  of  12°  to 
16°, 

The  number  of  degrees  multiplied  by  0-0225,  gives  the 
weight  of  lactic  acid  present  in  grams. 

Fat. — Owing  to  the  great  importance  of  the  estimation 
of  this  constituent,  numerous  methods  have  been  devised 
for  the  purpose.  These  may  be  divided  into  two  classes  ; 
i.e.,  those  giving  scientifically  accurate  results,  and  the 
more  rapid  processes  for  dealing  with  a  number  of 
samples  at  the  same  time,  which  give  results  of  sufficient 
accuracy  for  all  ordinary  work. 

Belonging  to  the  former  class  are  : — • 

(1)  Soxhlet's   method,    according   to   which   a   certain 
quantity  of  milk  is  mixed  with  certain  amounts  of  potash 
solution  and  water-saturated  ether,  the  ether  taking  up 
all  the  fat,  the  amount  of  which  is  estimated  by  taking  the 
specific  gravity  of  the  solution  at  a  certain  temperature. 

(2)  Wollny's  method,  in  which  the  refractive  index  of 
the  ethereal  fat  solution  is  determined  instead  of  the 
specific  gravity  as  in  the  previous  method. 

(3)  Extraction  methods,  in  which  a  known  quantity  of 
milk  is  soaked  into  porous  material  such  as  filter  paper 
(Adams),   or  granular   kaolin    (Wilson),   which   is  then 
extracted  with  ether  in  a  Soxhlet  or  similar  extraction 
apparatus,  the  fat  thus  obtained  being  weighed  after 
evaporation  of  the  ether. 

(4)  The  Rose-Gottlieb  Method. — As  this  is  one  of  the 
most  convenient  methods  for  the  determination  of  fat  in 
whole  milk,  skim  milk  or  cream,  besides  being  one  of  the 
most  trustworthy  available,  it  will  be  described  in  detail 
here.     In  passing,  mention  may  be  made  of  the  Werner- 
Schmidt,  and  the  Liebermann-Szekely  methods,  which 
resemble  the  Rose-Gottlieb  method  in  principle. 

13—2 


196      Industrial  Organic  Analysis 

The  determination  of  fat  in  whole  milk  or  skim  milk 
is  carried  out  as  follows  by  the  Rose-Gottlieb  method  :— 

10  c.c.  of  milk,  measured  from  a  burette  (equivalent  to 
10-27  grams)  or  about  10  grams  by  weight  of  milk,  are 
introduced  into  a  graduated  cylinder,  about  60  cm.  long, 
ij  to  2  cm.  in  diameter,  and  graduated  in  half  cubic 
centimetres  from  o  to  100  c.c.,  from  bottom  to  top.     I  c.c. 
of  aqueous  ammonia  of  specific  gravity  0*96  is  added 
from  a  pipette,  and  then  10  c.c.  of  95  per  cent,  alcohol. 
A  well-fitting  cork,  which  has  previously  been  soaked  in 
water,  is  inserted  into  the  mouth  of  the  tube,  and  the 
contents  are  gently  agitated  by  repeatedly  inverting  the 
tube  ;    the  casein  will  be  dissolved,  and  a  homogeneous 
liquid  will  be  obtained.     25  c.c.  of  pure  ether  are  then 
added,  and  the  contents  of  the  tube  mixed  as  before. 
Finally,  25  c.c.  of  petroleum  ether,  which  is  completely 
volatile   below  60°,    are   added,    the   contents   are  well 
mixed,  avoiding  violent  shaking,  and  allowed  to  stand 
at  rest  for  6  hours,  or  until  the  layers  have  completely 
separated  from  one  another.     The  object  of  the  addition 
of  the  petroleum  ether  is  to  precipitate  the  greater  part 
of  the  water  held  in  solution  by  the  ether.     All  the 
constituents  of  the  milk,  except  the  fat,  remain  in  the 
aqueous  alcoholic  layer  ;    it  now  remains  to  evaporate 
a  definite  proportion  of  the  ethereal  solution,  and  weigh 
the  residual  fat.     In  the  mouth  of  the  tube  is  inserted  a 
well-fitting  two-holed  cork  in  which  are  fitted  a  short 
tube  bent  at  right  angles  and  projecting  only  a  millimetre 
or  two  below  the  bottom  of  the  cork,  and  a  syphon  tube 
of  about  3  to  4  mm.   external  diameter,  bent  round 
sharply  so  that  both  limbs  are  nearly  parallel,  and  of  such 
a  length  that  all  the  ethereal  solution  can  be  drawn  off 
by  its  means  if  desired.     The  syphon  is  so  adjusted  that 
the  bottom  of  its  shorter  limb  comes  within  exactly 


Milk  and  Butter  197 

1-5  c.c.  of  the  surface  of  the  ammoniacal  alcohol  solution, 
the  cork  closing  the  mouth  of  the  graduated  tube.  The 
ethereal  fat  solution  is  started  running  by  blowing 
through  the  short  bent  tube,  and  caught  in  a  tared  flask 
of  about  100  to  150  c.c.  capacity.  After  again  gently 
blowing  through  the  shorter  tube,  in  order  to  drive  any  of 
the  ethereal  solution  which  may  remain  in  the  syphon 
into  the  flask,  the  ether  and  petroleum  ether  are  carefully 
evaporated  off  and  the  residual  fat  dried  at  105°  and 
weighed.  If  the  stated  measures  of  ether  and  petroleum 
ether  have  been  added,  and  loss  by  evaporation  has  been 
avoided,  then  the  1-5  c.c.  of  fat  solution  remaining  in  the 
tube  will  correspond,  nearly  enough,  to  0-27  gram  of  fat, 
so  that  if  10  c.c.  of  milk  were  taken,  the  fat  weighed  will 
correspond  to  exactly  10  grams  of  milk.  If,  on  the  other 
hand,  the  measurements  given  have  not  been  adhered  to, 
then  the  volume  of  the  solution  syphoned  off  must  be 
noted,  together  with  the  total  volume  of  the  fat  solution, 
and  a  calculation  made  accordingly.  In  the  case  of 
skim  milk,  this  latter  mode  of  procedure  must  be  adopted, 
or  the  whole  of  the  fat  may  be  weighed,  as  described 
below  for  the  determination  of  fat  in  cream  by  the 
present  method.  The  fat  weighed  should  be  perfectly 
clear  and  free  from  smell  of  petroleum  ether.  The 
drying  of  the  fat  should  not  be  unduly  prolonged,  or  gain 
in  weight  may  take  place  owing  to  oxidation  through 
exposure  to  air  at  the  temperature  of  the  oven. 

The  above  method  gives  accurate  results  with  whole 
milk,  skim  milk  and  buttermilk,  the  latter  being  treated 
in  the  same  way  as  skim  milk  in  respect  to  the  measure- 
ment of  the  fat  solution.  It  may  also  be  applied  to  the 
determination  of  fat  in  cream  as  follows  :  2  to  3  grams  of 
cream,  according  to  the  amount  of  fat  present,  are 
weighed  out  in  a  small  beaker  covered  with  a  watch 


198      Industrial  Organic  Analysis 

glass ;  the  bulk  of  the  cream  is  poured  into  the  graduated 
cylinder,  and  the  beaker,  with  its  cover,  is  re- weighed,  the 
amount  of  cream  taken  being  estimated  by  difference. 
After  diluting  the  cream  with  water  to  exactly  10  c.cv 
the  procedure  is  the  same  as  that  described  for  milk, 
with  the  exception  that  as  much  as  possible  of  the 
ethereal  fat  solution  is  removed  without  disturbing  the 
ammoniacal  alcoholic  layer,  after  which  the  remaining 
traces  of  fat  are  transferred  to  the  tared  flask  by  shaking 
out  with  two  successive  portions  of  50  c.c.  of  a  mixture 
of  ether  and  petroleum  ether  in  equal  parts  and  syphon- 
ing off  as  before.  The  whole  of  the  fat  present  in  the 
cream  is  thus  weighed  ;  this  mode  of  procedure  is  recom- 
mended for  dealing  with  cream,  as  small  quantities  of  fat 
are  apt  to  be  retained  by  the  ammoniacal  alcoholic 
solution  after  the  first  treatment  with  ether  and  petro- 
leum ether.  In  the  case  of  whole  milk,  skim  milk  or 
buttermilk,  however,  the  process  first  described  may 
safely  be  adopted. 

Modification  of  the  above  Method. — Eichloff  (Milch- 
wirtschaftliches  Centralblatt,  1910,  p.  114)  has  designed 
the  modified  form  of  apparatus  shown  in  Fig.  12 
for  carrying  out  the  determination  of  fat  accord- 
ing to  the  Rose-Gottlieb  process.  The  vase-shaped 
vessel  possesses  obvious  advantages  over  the  long 
graduated  tube  which  it  is  designed  to  replace,  being  of 
such  size  and  weight  that  it  can  conveniently  be  weighed 
on  the  balance  ;  the  milk  or  cream  may,  therefore,  be 
directly  weighed  out  in  it.  As  there  are  no  graduations, 
the  whole  of  the  fat  must  be  weighed ;  it  is,  however, 
only  necessary,  after  the  first  portion  of  the  fat  solution 
has  beeen  syphoned  off  as  completely  as  possible,  to  wash 
out  the  vessel  with  two  successive  portions  of  25  c.c.  of 
ether,  which  need  not  be  shaken  with  the  ammoniacal 


Milk  and  Butter 


199 


FIG.  12. — Apparatus  for  the  Rose-Gottlieb  Process. 


20O      Industrial  Organic  Analysis 


alcohol.  The  drying  of  the  fat  is 
considerably  facilitated  by  using 
the  specially  designed  beaker  flask 
shown  in  the  figure,  instead  of  an 
ordinary  flask.  After  the  solvent 
has  been  evaporated  off,  the  fat 
may  be^  completely  dried  within 
i  hour  ;  in  this  way,  errors  owing 
to  oxidation  are  avoided.  With 
the  modifications  just  mentioned, 
the  process  is  carried  out  in  the 
same  way  as  described  for  the 
original  Rose-Gottlieb  process. 

Rapid  Methods  for  the  Determi- 
nation of  Fat  in  Milk  and  Cream. 
— In  the  methods  which  are  most 
commonly  employed  in  practice, 
the  fat  is  separated  from  the  other 
constituents  of  the  milk  or  cream 
by  the  use  of  the  centrifuge,  and 
its  volume  read  off.  As  an 
example,  Gerber's  method  will  be 
outlined. 

Fig.  13  shows  one  form  of  butyro- 
meter  tube  which  may  be  used  for 
the  analysis  of  milk  by  this  pro- 
cess. Into  this  tube  are  introduced 
10  c.c.  of  concentrated  sulphuric 
acid  of  specific  gravity  1-825  at 
15°  C.,  and  ii  c.c.  of  the  milk, 
taken  from  a  well-mixed  sample, 
are  carefully  run  in  so  as  to  avoid 
complete  admixture  with  the  acid. 
Finally,  i  c.c.  of  amyl  alcohol  of 


Fat  in  Milk. 


Milk  and  Butter  201 

specific  gravity  0-815  at  15°  C.  and  boiling  point  between 
128°  and  130°  is  added.  The  acid  and  alcohol  which  are 
employed  to  facilitate  the  separation  of  the  fat  in  a 
clear  layer  are  generally  added  from  measures  which 
automatically  deliver  the  required  amounts.  A  well- 
fitting  long  rubber  stopper  is  inserted  in  the  butyrometer 
tube,  which  is  then  whirled  in  a  centrifuge  at  about 
1,000  revolutions  per  minute  for  3  to  4  minutes.  The 
tube  and  its  contents  are  brought  to  65°  in  a  water  bath, 
and  the  fat  percentage  is  directly  read  off  by  observing 
the  graduation  to  which  the  meniscus  of  the  fat  layer 
reaches,  the  lower  surface  having  been  brought  to  the 
zero  division  by  manipulating  the  rubber  stopper. 

Similar  tubes  may  be  obtained  for  the  analysis  of  skim 
milk  or  cream  by  this  method.  The  tubes  for  whole  milk 
are  usually  graduated  to  show  up  to  7. or  8  per  cent,  of 
fat,  while  those  for  cream  show  up  to  about  40  or  60  per 
cent,  of  fat,  using  either  5  grams  or  5  c.c.  of  cream.  The 
special  tubes,  measures,  and  all  other  requisites  for  the 
above  process,  may  be  obtained  from  most  dealers  in 
chemical  apparatus. 

A  large  number  of  determinations  may  be  rapidly 
carried  out  by  the  methods  involving  centrifuging,  the 
results  obtained  being  sufficiently  accurate  for  ordinary 
technical  and  commercial  work. 

Proteins. — The  method  of  estimation  of  the  total  pro- 
teins of  milk  by  treating  10  grams  of  milk  according  to 
the  Kjeldahl  process  for  estimating  nitrogen,  and  calcu- 
lating the  percentage  of  nitrogen  thus  found  to  proteins, 
by  multiplying  by  6-37,  does  not  give  very  reliable 
results,  as  the  whole  of  the  nitrogen  contained  in  the  milk 
is  not  present  as  proteins.  Better  results  are  obtained 
by  estimating  separately  the  casein  and  albumin  by  the 
methods  now  to  be  described. 


2O2      Industrial  Organic  Analysis 

Casein. — The  determination  of  this  constituent  must 
be  carried  out  on  fresh  or  nearly  fresh,  uncoagulated 
milk.  If  the  analysis  cannot  be  proceeded  with  during 
the  next  24  hours,  the  sample  should  be  preserved  as 
described  under  the  heading  of  "  Sampling,"  preferably 
by  keeping  it  in  a  refrigerator,  but  failing  this,  by  means 
of  potassium  dichromate. 

The  methods  by  which  casein  is  separated  from 
albumin  and  other  constituents  of  milk  depend  on  the 
fact  that  the  casein  is  salted  out,  i.e.,  precipitated,  by  the 
addition  of  solutions  of  certain  salts  such  as  alum  or 
magnesium  sulphate,  while  the  albumin  is  not  so  pre- 
cipitated, excepting  in  the  presence  of  free  acid.  After 
the  casein  has  been  salted  out  the  albumin  may  be  pre- 
cipitated by  the  addition  of  tannic  acid,  with  which  it 
forms,  like  most«of  the  proteins,  an  insoluble  compound  ; 
it  may  also  be  precipitated  from  the  saturated  mag- 
nesium sulphate  solution  by  boiling  with  a  little  dilute 
acetic  acid.  Both  the  casein  and  the  albumin  are 
estimated  as  nitrogen  by  the  Kjeldahl  process.  An 
alternative  method  consists  in  precipitating  the  casein 
by  adding  a  little  dilute  acetic  acid  to  the  diluted  milk 
and  estimating  the  albumin  in  the  nitrate  by  the 
Kjeldahl  process.  The  latter  method  is,  however,  not  so 
reliable  as  the  others  indicated  above,  as  the  casein  is 
not  always  quantitatively  precipitated  by  means  of  acid. 

The  American  Association  of  Official  Agricultural 
Chemists  recommend  the  following  method  for  the 
estimation  of  casein  in  milk  :— 

To  5  grams  of  milk  add  50  c.c.  of  a  solution  of 
magnesium  sulphate  saturated  at  45°,  and  heat  the 
mixture  to  45°  till  the  precipitate  separates  and  subsides, 
leaving  the  supernatant  liquor  clear.  Collect  the  pre- 
cipitate on  a  filter,  wash  two  or  three  times  with  a  solu- 


Milk  and  Butter  203 

tion  of  magnesium  sulphate,  as  used  above,  at  the  same 
temperature,  i.e.,  45°,  and,  after  drying  in  the  steam 
oven,  transfer  the  filter  and  precipitate  to  a  Kjeldahl 
digestion  flask  for  the  determination  of  the  nitrogen.1 
The  amount  of  nitrogen  found,  multiplied  by  6-25,  gives 
the  weight  of  casein  present  in  the  sample  analysed. 

An  alternative  method  for  precipitating  casein 
(Schlossman)  is  to  dilute  10  c.c.  of  milk  with  40  c.c.  of 
water,  and  to  add,  while  stirring,  i  c.c.  of  a  cold  saturated 
solution  of  alum,  when  the  casein  quickly  separates. 
If  the  supernatant  liquid  is  not  quite  clear  a  further 
quantity  of  alum  solution,  not  exceeding  0-5  c.c.,  is  added, 
drop  by  drop.  The  casein  is  then  filtered  off  and  esti- 
mated as  nitrogen  by  the  Kjeldahl  process. 

For  the  precipitation  of  the  casein  by  means  of  acid, 
the  following  method  is  recommended  by  the  American 
Official  Association  of  Agricultural  Chemists  : — 

10  grams  of  milk  contained  in  a  beaker  are  diluted 
with  90  c.c.  of  water  at  40°  to  42°,  and  1-5  c.c.  of  a 
10  per  cent,  solution  of  acetic  acid  in  water  is  added  at 
once.  The  mixture  is  stirred  with  a  glass  rod  and  allowed 
to  stand  for  5  minutes,  after  which  the  supernatant 
liquid  is  poured  off,  and  the  precipitate  washed  with  cold 
water  by  decantation.  The  casein  is  collected  on  a 
filter  and  estimated  as  nitrogen  by  the  Kjeldahl  process. 

Albumin. — The  filtrate  with  washings  obtained  by  any 
of  the  above  methods,  from  the  precipitated  casein,  is 
treated  with  10  c.c.  of  a  solution  of  4  grams  of  tannic  acid 
in  8  c.c.  of  25  per  cent,  acetic  acid,  which  has  been  made 
up  to  200  c.c.  with  40  to  50  per  cent,  alcohol ;  the  result- 
ing precipitate  is  filtered  off  after  settling,  washed,  dried 
and  analysed  by  the  Kjeldahl  process.  The  amount  of 

1  For  a  description  of  the  Kjeldahl  process,  see  Chapter  I.,  p.  22 


204      Industrial  Organic  Analysis 

nitrogen  found,  multiplied  by  6-34,  gives  the  amount  of 
albumin  (and  globulin)  present  in  the  milk. 

It  was  stated  above  that  albumin  is  only  precipitated 
from  a  saturated  solution  of  magnesium  sulphate  in 
presence  of  acid.  If  the  casein  has  been  precipitated  by 
means  of  this  salt,  as  in  the  first  of  the  methods  described 
above,  the  albumin  may  be  precipitated  by  neutralising 
the  filtrate  from  the  casein  with  dilute  sodium  hydroxide 
solution,  adding  0-3  c.c.  of  10  per  cent,  acetic  acid  and 
heating  in  boiling  water  for  10  to  15  minutes.  The 
precipitate  is  collected  on  a  filter  and  analysed  for 
nitrogen,  as  in  the  previous  method. 

It  is,  of  course,  necessary  that  the  filter  paper  used  for 
collecting  the  precipitated  proteins  should  be  either  free 
from  nitrogen  or  else  of  known  nitrogen  content,  as  it  is 
treated  together  with  the  proteins  in  the  Kjeldahl 
determination.  The  tannic  acid  used  should  also  be 
tested  to  make  sure  that  it  is  free  from  nitrogen. 

Milk  Sugar. — Two  methods  will  be  described  for  the 
determination  of  this  constituent  in  milk,  the  one  gravi- 
metric and  the  other  optical. 

Gravimetric  Method. — The  process  to  be  described,  due 
to  Allihn  and  Soxhlet,  depends  on  the  ability  of  lactose 
to  reduce  a  warm  alkaline  solution  of  a  cupric  salt 
(Fehling's  solution),  the  amount  of  the  insoluble  cuprous 
oxide  formed  under  given  conditions  being  taken  as  a 
measure  of  the  lactose  present.  It  will  be  seen  that  the 
same  principles  are  involved  as  in  the  method  for  the 
estimation  of  cane  sugar  described  in  Chapter  IV.,  p.  137, 
excepting  that  the  lactose  reduces  the  Fehling's  solution 
direct,  while  the  cane  sugar  itself  has  no  reducing  action, 
but  must  first  be  hydrolysed,  or  inverted,  by  heating  with 
acid.  The  amount  of  lactose  corresponding  to  a  given 
weight  of  copper  or  copper  oxide,  as  obtained  by  the 


Milk  and  Butter 


205 


method  to  be  described,  is  determined  by  reference  to 
the  accompanying  table,  as  the  amount  of  cuprous 
oxide  precipitated  is  not  strictly  proportional  to  the 
amount  of  lactose  present.  Before  the  actual  estimation 
of  the  lactose  in  milk  can  be  proceeded  with,  the  proteins 
must  be  removed  as  follows  (Ritthausen)  : — 

25  c.c.  of  milk  are  weighed  out  and  diluted  with  400  c.c. 
of  water  in  a  500  c.c.  flask.  The  proteins  are  precipitated 
by  adding  10  c.c.  of  Fehling's  copper  solution  (containing 
69-28  grams  of  CuSO4,  5H2O  per  litre,  not  the  alkaline 
tartrate  solution),  and  then  sufficient  decinormal  sodium 
hydroxide  solution  to  give  a  neutral  or  feebly  acid 

SOXHLET'S  TABLE  FOR  FINDING  THE  WEIGHT  OF  LACTOSE  CORRE- 
SPONDING   TO    A    GIVEN    WEIGHT     OF    COPPER     REDUCED 

ACCORDING   TO   ALLIHN    AND  SOXHLET'S  METHOD    OF   DETER- 
MINING LACTOSE. 


Cu, 
milli- 
grams. 

Lactose, 
milli- 
grams. 

Cu, 
milli- 
grams. 

Lactose, 
milli- 
grams. 

Cu, 
milli- 
grams. 

Lactose, 
milli- 
grams. 

140 

ioi-3 

240 

176-9 

34° 

255-8 

145 

105-1 

245 

180-9 

345 

259-8 

150 

108-8 

250 

184-8 

350 

263-9 

155 

112-6 

255 

188-7 

355 

268-0 

1  60 

116-4 

260 

192-6 

360 

272-1 

165 

I2O-2 

265 

196-4 

365 

276-3 

170 

123-9 

270 

200-3 

37° 

280-5 

175 

I2y8 

275 

204-3 

375 

284-8 

1  80 

I3I-6 

280 

208-3 

380 

289-1 

I85 

I35-4 

285 

212-3 

385 

293-3 

IQO 

139-2 

290 

2I6-3 

390 

297-7 

195 

M3'1 

295 

22O-3 

395 

302-0 

2OO 

146-9 

300 

224-4 

400 

306-3 

205 

I50-7 

305 

228-3 

210 

154-4 

310 

232-I 

215 

158-2 

315 

236-0 

220 

I6I-9 

320 

239-9 

225 

I65-6 

325 

243-8 

230 

169-4 

330 

247-7 

235 

I73-I 

335 

25I-7 

Factor  for  convert- 
ing CuO  to  Cu  = 
0-7989. 


206      Industrial  Organic  Analysis 

reaction.  Usually,  from  6-5  to  7-5  c.c.  of  the  alkali 
solution  will  be  sufficient ;  in  no  case  should  enough  be 
added  to  produce  an  alkaline  reaction.  20  c.c.  of  a  cold 
saturated  solution  of  sodium  fluoride  are  then  added  in 
order  to  remove  the  dissolved  calcium  salts,  the  presence 
of  which  tends  to  influence  the  results  of  the  sugar 
determination.  After  allowing  to  settle  for  half  an  hour 
the  flask  is  filled  to  the  500  c.c.  graduation,  and  the  clear 
solution  filtered  through  a  dry  filter.  In  a  deep  porce- 
lain dish,  50  c.c.  of  Fehling's  solution  (made  by  mixing 
25  c.c.  of  the  copper  sulphate  solution,  as  used  previously, 
with  25  c.c.  of  a  solution  containing  250  grams  of 
potassium  hydroxide  and  350  grams  of  Rochelle  salt  per 
litre)  are  heated  to  boiling,  and  100  c.c.  of  the  filtrate 
containing  the  lactose,  obtained  as  just  described,  are 
added.  After  stirring,  the  mixture  is  boiled  for  exactly 
6  minutes,  and  the  precipitate  formed  is  filtered  off  on  a 
Gooch  crucible,  washed  with  water,  alcohol  and  ether, 
ignited  and  weighed  as  CuO,  as  described  on  p.  140. 

Polarimetric  Method. — In  this  method  advantage  is 
taken  of  the  optical  activity  of  lactose.  A  clear  whey, 
containing  the  lactose,  may  be  prepared  for  polarimetric 
examination  by  the  following  process,  due  to  Wiley  :— 

60  c.c.  of  milk  are  treated  with  10  c.c.  of  a  solution  of 
mercuric  nitrate,  prepared  by  dissolving  mercury  in 
twice  its  weight  of  nitric  acid  of  specific  gravity  1-42,  and 
diluting  the  resulting  solution  with  4  times  its  volume  of 
water.  The  mixture  of  milk  and  clearing  solution  is 
made  up  to  100  c.c.,  filtered  through  a  dry  filter  and 
transferred  to  a  2  or  4  decimetre  tube  for  the  determina- 
tion of  optical  activity  at  17-5°  C.  The  specific  rotation 
of  the  hydrated  lactose,  C12H22OnH2O,  is  [a]D  +  52-5°. 
Hence  if  C  be  the  concentration  of  the  lactose  in  the  solu- 
tion under  observation,  expressed  in  grams  per  100  c.c., 


Milk  and  Butter  207 

C  =  -          y  ,  where  D  is  the  angular  rotation  observed, 
52'5  X  1- 

and  L  the  length  of  the  tube  in  decimetres.  The  volume 
of  the  solution,  after  clearing  and  making  up  to  100  c.c., 
is  100  c.c.  minus  the  volume  of  the  precipitate,  containing 
the  fat  and  proteins,  which  is  allowed  for  as  follows  :  the 
weight  of  fat  in  grams  contained  in  the  milk  taken, 
multiplied  by  1-075,  and  the  weight  of  proteins,  multi- 
plied by  0-8,  subtracted  from  100,  gives  the  volume  of  the 
solution  in  cubic  centimetres.  The  weight  of  hydrated 
milk  sugar  in  60  c.c.  of  milk  is  thus  found.  Dividing  by 
the  specific  gravity  of  the  milk  the  weight  in  60  grams  of 
milk  is  arrived  at,  and  further,  multiplying  by  0-95  will 
give  the  weight  of  anhydrous  milk  sugar  present. 

The  percentage  of  anhydrous  milk  sugar  in  the  milk 
may  thus  be  calculated  by  the  following  formula  :  — 

Percentage  of  anhydrous  lactose 


=  C  x   I0°  ~  (i'Q75F  +  Q-8P)       IPO  X  0*95 
Q  sp.  gr. 

where  Q  =  the  number  of  cubic  centimetres  of  milk 
taken,  F  =  the  weight  of  fat  in  the  quantity  Q,  P  =  the 
weight  of  proteins  in  the  quantity  Q. 

Total  Solids.  —  About  20  grams  of  milk  are  weighed  out 
in  a  flat  dish,  evaporated  to  dryness  on  the  water  bath, 
and  the  residue  is  dried  at  105°  till  constant  in  weight. 
In  order  to  prevent  the  formation  of  a  skin  on  the  surface 
of  the  milk  a  few  drops  of  alcohol  or  acetic  acid  may  be 
added  ;  the  process  of  evaporation  will  thereby  be 
facilitated. 

The  total  solids  in  milk  may  also  be  calculated  from 
the  specific  gravity,  by  Fleischmann's  formula,  as 
described  under  the  heading  of  the  "  Detection  of  the 
Adulteration  of  Milk,"  where  the  interpretation  of  the 
results  of  the  determination  will  be  discussed.  The 


208      Industrial  Organic  Analysis 

last  mentioned  method  is  usually  employed  in  practice, 
on  account  of  its  rapidity. 

Ash. — 20  grams  of  milk  are  weighed  into  a  platinum 
dish  and  evaporated  to  dry  ness  on  the  water  bath,  with 
the  addition  of  a  few  drops  of  acetic  acid  or  alcohol  to 
prevent  the  formation  of  a  skin.  The  residue  is  car- 
bonised at  a  clear  red  heat,  cooled  and  extracted  with 
water.  The  aqueous  extract  is  filtered,  taking  care  that 
as  little  as  possible  of  the  carbonaceous  matter  is  brought 
on  to  the  filter.  The  remaining  coal,  which  has  now 
been  freed  from  most  of  the  mineral  matter,  is  burnt  to 
ash,  after  which  the  aqueous  extract  containing  the 
mineral  matter  is  returned  to  the  dish  and  evaporated 
to  dryness.  The  whole  is  then  heated  at  a  low  red  heat 
until  a  perfectly  white  ash  is  obtained,  cooled  and 
weighed. 

The  above  procedure  must  be  adopted  if  accurate 
results  are  desired.  If  the  total  solids  are  directly 
incinerated  undue  loss  will  occur.  As  it  is,  a  slight  loss 
will  always  occur  owing  to  the  decomposition  of  the 
citrates,  loss  of  water  by  the  mono-  and  di-phosphates 
and  volatilisation  of  the  chlorides. 

If  more  than  075  per  cent,  of  ash  is  found,  the  milk  is 
probably  abnormal.  According  to  Vieth,  the  proportion 
of  lactose  to  proteins,  to  ash,  in  normal  milk,  should  be 
as  13  to  9  to  2.  In  abnormal  milks,  the  ash  is  often  high 
and  the  milk  sugar  content  low.  It  will  easily  be  under- 
stood how  such  a  state  may  be  brought  about  by  the 
addition  of  water. 

Other  Constituents. — For  the  determination  of  the  other 
constituents  of  milk,  such  as  citric  acid  and  lecithin,  see 
the  works  mentioned  at  the  end  of  this  chapter. 

THE  DETECTION  OF  THE  ADULTERATION  OF  MILK. 

The  most  commonly  practised  methods  of  adulterating 
milk  are  as  follow  :  (i)  Addition  of  water,  (2)  removal 


Milk  and  Butter  209 

of  cream,  (3)  simultaneous  addition  of  water  and 
removal  of  cream.  While  it  is  of  the  greatest  importance 
that  such  an  important  foodstuff  as  milk  should  reach 
the  consumer  in  an  unadulterated  condition,  the  addition 
of  water  is  especially  to  be  deprecated  on  account  of  the 
risk  attaching  thereto  of  introducing  pathogenic  germs 
such  as,  for  example,  those  of  typhus,  diphtheria  or 
cholera,  all  of  which  are  capable  of  multiplying  rapidly 
in  milk. 

(i)  Addition  of  Water.  Nitrate  Reaction. — This  reac- 
tion, which  is  often  prescribed  for  the  detection  of  added 
water,  is  based  on  the  fact  that  normal  milk  should 
contain  no  nitrates,  and  on  the  assumption  that  most 
well  and  tap  waters  contain  small  quantities  of  nitrates. 
The  following  test  for  nitrates  is  accordingly  applied 
with  the  object  of  detecting  added  water  :  to  10  c.c.  of 
milk  add  i  drop  of  formaldehyde  solution  and  about 
10  c.c.  of  pure  sulphuric  acid  of  specific  gravity  1*815. 
In  the  presence  of  nitrates,  a  bluish  violet  coloration  is 
developed,  which  also  depends  on  the  presence  of  milk 
proteins. 

In  addition  to  the  fact  that  all  waters  do  not  contain 
nitrates,  it  should  be  noted  that  the  washing  out  of  the 
tanks  and  other  vessels  used  for  storing  the  milk  with 
water  which  contains  appreciable  quantities  of  nitrates 
may  give  rise  to  a  perceptible  nitrate  reaction,  even 
though  the  milk  be  perfectly  genuine.  The  indications 
afforded  by  this  test  therefore  require  confirmation  by 
other  means,  as  indicated  below. 

Specific  Gravity,  Fat  Percentage  and  Solids  less  Fat. — 
The  tables  on  p.  183  show  the  extreme  variations  in  the 
composition  of  normal  milk  from  many  different  sources. 
The  percentage  of  solids  less  fat  is  subject  to  least  varia- 
tion, and  is,  therefore,  generally  used  as  a  basis  for  the 
detection  of  added  water,  and  the  calculation  of  the 
extent  of  the  adulteration.  The  addition  of  water  to 
milk  will  obviously  tend  to  lower  the  percentage  of  solids 

1.0. A.  14 


210      Industrial  Organic  Analysis 

less  fat,  and  also  the  specific  gravity  and  the  percentage 
of  fat.  As  was  stated  above,  in  the  United  Kingdom  it 
is  assumed,  for  purposes  of  the  Sale  of  Food  and  Drugs 
Act,  that  genuine  milk  should  contain  at  least  3-0  per 
cent,  of  fat,  and  8-5  per  cent,  of  solids  less  fat.  Mixed 
milk,  in  which  the  percentage  of  solids  less  fat  falls  below 
8-4,  is  almost  certain  to  have  been  adulterated  by  the 
addition  of  water,  while  a  specific  gravity  of  1-028  will 
render  the  milk  liable  to  suspicion  ;  a  lower  value  will 
indicate  that  adulteration  with  water  is  practically 
certain  to  have  taken  place.  The  specific  gravity  of 
milk  is  lowered  by  approximately  3°,  i.e.,  3  places  in  the 
third  decimal  place,  for  each  10  per  cent,  of  added  water. 
In  order  to  detect  the  addition  of  water  to  milk  and  to 
calculate  the  probable  extent  of  the  adulteration,  it  is 
usual  to  determine  the  fat  percentage  and  the  specific 
gravity,  and  from  these  to  calculate  the  percentage  of 
solids  less  fat  by  the  use  of  Fleischmann's  formula  : — 

_    ,       rf    loos  —  100 
A  =  i'2  X  F  -f-  2-665 ~ 

and  S  =  A  —  F, 

where  A  is  the  percentage  of  total  solids,  including  the 
fat,  F  is  the  percentage  of  fat,  s  the  specific  gravity  of  the 
milk  at  15°  C.,  and  S  the  percentage  of  solids  less  fat. 

This  formula  was  elaborated  by  Fleischmann  from  the 
results  of  numerous  analyses ;  it  depends  on  the  relative 
constancy  in  the  composition  of  the  solids  less  fat  in 
normal  milks  from  different  sources  ;  naturally,  it  will 
only  hold  good  for  milks  which  are  normal  in  this  respect, 
i.e.,  those  in  which  the  proteins,  ash  and  lactose  are 

E resent  approximately  in  the  proportion  of  9  to  2  to  13. 
n  general,  however,  this  proportion  may  be  assumed  to 
hold  good  for  fresh  milk.  Instead  of  applying  the  above 
formula  each  time  the  solids  less  fat  are  to  be  determined, 
it  is  usual  to  refer  to  a  table  containing  data  based  on 
this  formula.  A  method  for  the  calculation  of  the 
probable  extent  of  the  adulteration  with  water  will  be 
given  below,  after  the  other  methods  of  adulteration  have 
been  discussed. 

(2)  Removal  of  Cream. — While  the  direct  addition  of 


Milk  and  Butter  211 

water  to  milk  gives  rise  to,  and  may  be  detected  by,  a 
lowering  of  the  specific  gravity,  the  removal  of  cream  has 
the  opposite  effect,  as  the  fat  is  the  lightest  constituent 
of  the  milk.  The  rise  produced  in  the  specific  gravity  is, 
however,  too  small  to  be  of  any  use  as  an  indication  of 
the  amount  of  fat  removed,  the  best  method  for  detecting 
a  fraud  of  this  nature  being  that  which  would  naturally 
suggest  itself,  i.e.,  the  estimation  of  the  fat  itself.  If  the 
fat  falls  below  3  per  cent.,  the  milk  is  liable  to  suspicion, 
and  in  a  law  case  it  would  rest  with  the  supplier  to  prove 
that  it  was  genuine.  From  the  fat  percentage  and  the 
specific  gravity,  the  total  solids  may  be  calculated  by 
Fleischmann's  formula,  as  described  above  ;  the  per- 
centage of  fat  in  solids  may  then  be  arrived  at  by  a 
further  simple  calculation  ;  if  this  is  found  to  be  under 
20 -o,  then  adulteration  of  the  milk,  either  by  simple 
removal  of  cream  or  addition  of  skim  milk,  is  practically 
certain  to  have  taken  place.  Further,  if  skim  milk  has 
been  added,  the  solids  less  fat  will  be  higher  than  in  the 
unadulterated  milk.  To  sum  up,  a  sample  of  milk  from 
which  cream  has  been  removed  or  to  which  skim  milk 
has  been  added  will  show  a  normal  or,  perhaps,  slightly 
increased  specific  gravity  and  percentage  of  solids  less 
fat,  and  a  low  percentage  of  fat  in  solids.  On  the  other 
hand,  the  addition  of  water  will  be  detected  by  a  low 
specific  gravity  and  percentage  of  solids  less  fat ;  while 
the  percentage  of  fat  in  the  milk  may  be  lowered  per- 
ceptibly, the  content  of  fat  in  solids  will  not  be  affected. 
(3)  Simultaneous  Removal  of  Cream  and  Addition  of 
Water  or  Skim  Milk. — An  adulteration  of  this  nature  may 
be  carried  out  in  such  a  way  that  the  specific  gravity  of 
the  sample  remains  normal,  the  lowering  which  would 
be  occasioned  by  the  addition  of  water  alone  being 
counterbalanced  by  the  rise  owing  to  the  abstraction  of 
fat.  In  any  case,  however,  the  fraud  will  be  detected  by 
a  lowering  in  the  percentage  of  fat  in  the  milk,  and 
especially  by  a  low  percentage  of  fat  in  solids.  If  water 
has  been  added,  the  percentage  of  solids  less  fat  in  the 
milk  will  be  lowered.  If,  on  the  other  hand,  skim  milk 
has  been  added,  the  latter  figure  may  be  slightly  raised  ; 
in  this  case  the  nature  of  the  adulteration  will  be  prac- 

14—2 


212      Industrial  Organic  Analysis 

tically  identical  with  that  discussed  under  the  previous 
heading.  The  percentage  of  fat  in  solids  will  naturally 
be  lowered  to  a  greater  extent  by  the  addition  of  skim 
milk  than  by  the  addition  of  an  equal  proportion  of  water, 
supposing  equal  amounts  of  fat  to  have  been  removed  in 
both  cases,  for,  in  the  first  case,  not  only  is  fat  abstracted, 
but  a  further  quantity  of  non-fatty  solids  is  added  with 
the  skim  milk,  while  in  the  second  case  the  fat  is  simply 
reduced  without  appreciable  amounts  of  non-fatty  solids 
being  introduced  with  the  water. 

The  foregoing  discussion  of  the  methods  for  the  detec- 
tion of  the  adulteration  of  whole  milk  may  be  briefly 
summarised  as  follows  : — 

(1)  Addition    of   Water. — Detected    by    low    specific 
gravity  and  solids  less  fat:  fat  in  milk,   more  or  less 
lowered ;  fat  in  solids,  normal. 

(2)  Removal  of  Cream  or  Addition  of  Skim  Milk. — 
Detected  chiefly  by  low  fat  in  solids :  fat  in  milk,  lowered ; 
specific  gravity  and   solids  less  fat,  normal  or   slightly 
raised. 

(3)  Simultaneous  Removal  of  Cream  and  Addition  of 
Water. — Detected  chiefly  by  low  fat  in  solids  and  low 
solids  less  fat:  specific  gravity  may  be  normal  or  lowered; 
fat  in  milk,  lowered. 

Addition  of  Water  to  Skim  Milk. — The  adulteration  of 
skim  milk  with  water  may  be  detected,  as  in  the  case  of 
whole  milk,  by  the  lowering  of  the  solids  less  fat,  the 
lower  limit  for  which,  in  the  case  of  skim  milk,  is  generally 
placed  at  9-0. 

The  methods  for  approximately  calculating  the 
probable  extent  of  the  adulteration  of  milk,  as  described 
above,  will  now  be  outlined. 

(i)  Addition  of  Water. — For  reasons  already  stated, 
the  calculation  is  usually  based  on  the  solids  less  fat. 
The  amount  of  water  added  to  100  parts  of  milk  may  be 
calculated  from  the  following  equation  :- — 

v       100  . 

X  =  — (r-  rj, 

where  X  is  the  amount  of  water  added  to  100  parts  of 
milk,  and  r  the  standard  percentage  of  solids  less  fat 


Milk  and  Butter  213 

adopted  for  comparison.  As  was  mentioned  above,  the 
lower  limit  for  the  solids  less  fat  in  milk  is  usually  placed 
at  8-5.  In  order  to  arrive  at  the  probable  extent  of  the 
adulteration,  most  analysts  take  the  standard  at  8-9  per 
cent.,  thus  assuming  that  the  milk  was  normal  in  this 
respect  before  the  water  was  added  ;  if,  however,  r  is 
taken  at  8-5,  a  minimum  figure  for  the  amount  of  added 
water  will  be  obtained.  In  cases  of  dispute,  the  solids 
less  fat  are  sometimes  determined  in  a  sample  taken  under 
proper  supervision  from  the  same  source  as  the  suspected 
sample,  and  the  value  found  substituted  for  r  in  the 
above  formula.  rv  is  in  all  cases  the  percentage  of  solids 
less  fat  found  in  the  suspected  sample. 

(2)  Removal  of  Cream.  —  The  amount  of  fat  abstracted 
may  be  calculated  from  the  following  formula  :  — 

Y  = 


where  Y  is  the  percentage  of  the  original  amount  of  fat 
removed,  fl  is  the  percentage  of  fat  in  the  original 
milk,  i.e.,  the  standard  value,  and  /2  the  percentage 
of  fat  in  the  suspected  sample.  f1  may  be  the  per- 
centage of  fat  in  a  sample  taken  under  proper  super- 
vision from  the  same  source  as  the  suspected  sample,  or, 
failing  this,  fl  should  be  taken  at  3-0,  the  minimum  fixed 
for  the  purposes  of  the  Sale  of  Food  and  Drugs  Act. 
The  probable  amount  of  fat  removed  will,  however,  in 
the  majority  of  cases,  be  more  nearly  arrived  at  by 
taking  fa  at  a  somewhat  higher  value,  say  3-4. 

(3)  Simultaneous  Removal  of  Cream  and  Addition  of 
Water.  —  An  approximate  idea  of  the  extent  of  the 
adulteration  may  be  arrived  at  by  applying  the  following 
two  formulae  :  — 

X  =  -  (wl  -  w)  and  Y  =  100  (i  -  ^,  - 
'i  J*\ 

where  X  is  the  amount  of  water  added  to  100  parts  of 
milk,  r  the  percentage  of  solids  less  fat  taken  as  the 
standard  value  for  comparison  (see  above),  rv  the  per- 
centage of  solids  less  fat  in  the  suspected  sample,  w  the 
percentage  of  water  in  the  original  milk  (obtained  by 


214      Industrial  Organic  Analysis 

subtracting  the  percentage  of  total  solids,  including  fat, 
from  100),  wl  the  percentage  of  water  in  the  suspected 
sample,  Y  the  amount  of  fat  taken  from  100  parts  of 
milk,  /  the  percentage  of  fat  in  the  original  milk,  and  fl 
the  percentage  of  fat  in  the  suspected  sample.  In  the 
absence  of  a  genuine  sample  from  the  same  source  as  the 
suspected  sample,  for  comparison,  the  total  solids  for 
calculation  of  the  amount  of  water  in  the  original  milk 
should  be  taken  at  3-0  +  8-5,  i.e.,  11*5,  though  a  more 
probable  value  would  probably  be  obtained  -in  the 
majority  of  cases  by  placing  it  at,  say,  3-5  +  8-9,  i.e., 
12-4.  As  was  pointed  out  above,  by  taking  the  minimum 
values  for  the  standard  percentages  of  fat  and  solids  less 
fat,  minimum  estimates  of  the  extent  of  the  adulteration 
will  be  obtained,  thus,  to  a  certain  extent,  giving  the 
supplier  of  the  suspected  sample  the  benefit  of  such 
doubt  as  may  arise  from  the  limitations  of  the  analytical 
methods  available. 

THE  HEAT  STERILISING  AND  PASTEURISING  OF  MILK. 

It  is  a  well-known  fact  that  all  micro-organisms  are 
destroyed  when  exposed  to  a  sufficiently  high  tempera- 
ture. On  this  principle  it  is  possible  to  improve  the 
keeping  powers  of  milk  by  temporarily  heating  it  to  a 
suitable  temperature.  It  has  been  found  that  milk  can 
only  be  completely  sterilised  by  heating  it  for  at  least 
half  an  hour  to  a  temperature  of  about  120°  by  means 
of  steam  under  pressure ;  under  this  treatment,  however, 
it  undergoes  such  a  change  in  taste  as  to  render  it  useless 
as  food,  and  in  practice  it  has  been  found  convenient  to 
employ  lower  temperatures,  at  which  only  a  slight 
"  cooked "  flavour  is  produced  in  the  milk.  It  was 
shown  by  Pasteur  that  temperatures  lower  than  100° 
could  be  employed  with  success  in  destroying  most  of 
the  micro-organisms  of  milk,  and  Weigmann  has  shown 
that  momentary  heating  to  85°  is  sufficient  to  destroy 
the  most  resistant  of  the  pathogenic  bacteria  which  are 
likely  to  occur  in  milk.  Some  bacteria,  yeasts  and 
moulds,  however,  produce  highly  resistant  spores  which 
are  only  destroyed  by  a  somewhat  more  prolonged 


Milk  and  Butter  215 

heating  at  higher  temperatures,  and  consequently 
survive  the  pasteurising  process  in  which  the  milk  is 
heated  to  temperatures  ranging  from  about  70°  to  85°, 
either  momentarily,  or,  at  most,  for  a  few  minutes. 
The  heating  is  carried  out  either  in  glass  vessels  immersed 
in  a  water  bath  at  the  desired  temperature,  or  in  specially 
constructed  continuously  working  pasteurising  appli- 
ances. After  heating,  the  milk  is  quickly  cooled,  in 
order  to  check,  as  far  as  possible,  the  development 
and  growth  of  the  surviving  organisms,  and  preserved 
so  as  to  avoid  infection  from  the  air  or  other  sources. 

Pasteurised  milk  will  generally  keep  unchanged  some 
24  to  26  hours  longer  than  unpasteurised  milk,  while 
sterilised  milk  will  keep  for  much  longer,  often  for  months, 
without  any  change  being  noticeable,  provided,  of  course,, 
that  it  is  preserved  so  that  risk  of  infection  from  outside 
sources  is  avoided.     It  should,  however,  be  noted  that 
if  pasteurised  milk  is  kept  too  long,  the  changes  which  it 
will  undergo,  owing  to  the  action  of  micro-organisms, 
will  have  much  more  serious  consequences  than  in  the 
case  of  unpasteurised  milk.     The  reasons  for  this  are  as 
follows  :  in  ordinary  milk,  kept  at  the  ordinary  tempera- 
ture,   the    relatively    harmless    lactic    acid    producing 
bacteria  develop  at  a  far  more  rapid  rate  than  all  the 
other  types  of  organisms  present,  and  thus  effectively 
keep  the  latter  in  check.     In  this  first  stage  of  decom- 
position,   the    milk   is    protected   against    putrefactive 
decomposition  leading  to  the  production  of  poisonous 
substances,  while  some  time  will  always  elapse  before 
sufficient  free  lactic  acid  has  been  produced  to  coagulate 
the  milk  ;    even  when  this  has  occurred,  however,  the 
milk  will  not  contain  any  poisonous  substances  resulting 
from  the  souring  process,   and  cannot  be  said  to  be 
unfit   for   food.     The  lactic   acid  produced   effectively 
checks   the   development   of   the   protein   hydrolysing, 
putrefactive    bacteria    which    only    become    active    in 
neutral  or  feebly  alkaline  media,  and  it  is  only  after  the 
souring  process  has  come  to  an  end  (i.e.,  when  the  milk 
contains  about  0-8  per  cent,  of  lactic  acid),  and  the  lactic 
acid  has  been  consumed  by  the  yeasts  or  moulds  which 
usually  follow  after  the  lactic  acid  producing  organisms 


216      Industrial  Organic  Analysis 

have  ceased  to  develop,  that  the  putrefactive  organisms 
are  able  to  develop  and  act  on  the  milk. 

When  the  milk  has  been  pasteurised,  the  conditions 
are  entirely  altered ;  the  lactic  acid  organisms  are 
destroyed  and  leave  no  spores  which  might  survive  the 
elevated  temperature.  Many  putrefactive  organisms 
(e.g.,  B.  subtilis,  or  the  hay  bacillus),  on  the  other  hand, 
form  highly  resisting  spores  which  survive  the  pasteuris- 
ing process  and,  on  cooling,  are  able  to  develop. in  the 
absence  of  lactic  acid  bacteria.  The  first  stage  in  the 
decomposition  of  pasteurised  milk  is,  therefore,  usually 
due  to  the  development  of  putrefactive  organisms  which 
hydrolyse  the  proteins  of  the  milk  into  the  bitter 
poisonous  peptones,  and  further  into  polypeptides  and 
amino  acids.  In  spite  of  what  has  just  been  said,  how- 
ever, pasteurising  has  proved  to  be  a  useful  method  of 
preserving  milk,  and  should  certainly  be  insisted  on 
when  the  milk  is  suspected  to  have  been  derived  from 
contaminated  sources.  In  modern  dairies,  where  milk 
from* many  different  sources  is  mixed  for  skimming, 
pasteurising  forms  a  regular  feature  in  the  process  of 
butter  and  margarine  making. 

The  effect  of  pasteurising  on  milk  may  be  readily 
shown  as  follows :  Two  clean  flasks  of  about  200  c.c. 
capacity  are  plugged  at  the  mouth  with  cotton-wool, 
and  sterilised  by  heating  to  about  140°  in  an  air  oven, 
for  about  20  minutes.  They  are  then  allowed  to  cool, 
without  removing  the  cotton-wool  plugs,  and  nearly 
filled  with  fresh  milk,  the  plugs  being  replaced  as  soon 
as  possible  after  the  introduction  of  the  milk.  One  of 
the  flasks  is  heated  in  boiling  water  for  about  20  minutes, 
and  then  cooled  in  running  water.  The  two  flasks  are 
allowed  to  stand  together  at  a  temperature  of  about 
20°  to  25°,  and  the  contents  examined  from  time  to 
time  ;  within  about  24  hours,  the  unheated  milk  should 
have  acquired  a  sour  taste  and  smell,  and  should  sooner 
or  later  become  coagulated  ;  the  unheated  milk,  on  the 


Milk  and  Butter  217 

other  hand,  should  remain  unchanged  for  a  much  longer 
time,  not  developing  acidity,  and  differing  only  from 
new  milk  in  having  a  slightly  "  cooked  "  flavour.  After 
the  unheated  milk  has  become  coagulated,  the  acidity 
of  the  two  samples  may  be  compared  by  titrating  a 
small  portion  of  each  with  standard  sodium  hydroxide 
solution,  using  phenol  phthalein  as  indicator.  On 
further  standing,  however,  it  will  be  found  that  a  bitter 
peptone-like  taste  and  smell  of  decay  will  develop  in  the 
heated  sample  quicker  than  in  the  unheated  sample. 

Test  for  Pasteurising,  (a)  Milk  and  Cream. — In  cases 
where  pasteurising  is  obligatory,  it  is  desirable  to  have 
some  means  of  ascertaining  whether  the  milk  or  cream 
for  butter  making  has  been  sufficiently  heated.  A  test 
commonly  employed  is  that  devised  by  Storch. 
Unheated  milk  decomposes  hydrogen  peroxide  with  the 
formation  of  oxygen  and  water,  the  active  substance 
presumably  being  an  enzyme  present  in  the  milk  to 
which  the  name  "  catalase  "  has  been  given.  When 
the  milk  has  been  heated  to  80°,  it  will  be  found  to 
have  lost  the  power  of  decomposing  hydrogen  peroxide, 
owing  to  the  destruction  of  the  catalytic  substance 
during  the  heating. 

The  test  is  carried  out  as  follows  :  To  5  c.c.  of  the 
sample  in  a  test  tube  add  i  drop  of  a  2  per  cent,  aqueous 
solution  of  hydrogen  peroxide,  shake,  then  add  2 
drops  of  a  2  per  cent,  aqueous  solution  of  para- 
phenylene  diamine,  and  mix  again  by  shaking.  If  the 
milk  has  not  been  heated,  the  nascent  oxygen  from  the 
hydrogen  peroxide  will  act  on  the  paraphenylene 
diamine,  giving  rise  to  a  blue  coloration  within  a  few 
seconds.  If  the  milk  has  been  heated  above  80°,  no 
colour  will  be  developed  in  30  seconds,  while  if  it  has 
only  been  heated  to  79°  to  80°  a  greyish  blue  colour 


218      Industrial  Organic  Analysis 

will  form  in  about  30  seconds.  This  test  may  easily  be 
verified  with  heated  and  unheated  milk. 

(b)  Butter. — The  test  just  described  may  also  be 
applied  to  butter,  with  a  view  to  ascertaining  whether  it 
has  been  made  from  properly  pasteurised  cream. 

25  grams  of  the  butter  are  melted  in  a  small  beaker  at 
a  temperature  not  above  60°.  When .  the  fat  has 
separated  clear  from  the  aqueous  layer,  it  is  syphoned 
off.  The  aqueous  serum,  which  is  turbid  and  contains 
a  white  sediment  of  casein,  is  diluted  with  an  equal  bulk 
of  water,  poured  into  a  test  tube  and  tested  with  hydro- 
gen peroxide  and  paraphenylene  diamine  solutions,  as 
directed  above. 

THE  BACTERIAL  CONTENTS  OF  MILK. 

Reductase  Test. — The  test  which  is  now  to  be  described 
affords  a  means  of  ascertaining  the  relative  number  of 
micro-organisms  present  in  milk.  It  has  already  been 
pointed  out  that  the  alteration  in  flavour,  and  decom- 
position of  milk,  are  due  to  the  action  of  the  numerous 
species  of  micro-organisms  which  grow  and  reproduce 
in  it.  It  is  obvious  that  the  greater  the  number  of 
micro-organisms  present,  the  quicker  will  be  the  decom- 
position of  the  milk.  Milk  as  it  leaves  the  udder  of  a 
healthy  cow  contains  relatively  few  organisms,  all  of 
which  are  generally  of  a  harmless  nature.  From  this 
moment  onwards,  however,  it  is  liable  to  become 
infected  by  organisms  from  the  air,  or  any  dirt,  fodder, 
straw,  etc.,  with  which  it  may  come  into  contact. 
Unless,  therefore,  the  greatest  care  and  cleanliness  be 
observed  in  the  treatment  of  the  milk,  its  keeping 
powers  will  be  materially  reduced. 

The  reductase  test  depends  on  the  ability  of  milk  to 
reduce  methylene  blue  to  a  colourless  substance  ;  the 
reduction  is  not  a  function  of  the  milk  itself,  but  of  the 
micro-organisms  which  it  contains,  or  rather  of  enzymes 
produced  by  the  latter.  The  greater  the  number  of 


Milk  and  Butter  219 

micro-organisms  present  the  shorter  will  be  the  time 
required  for  the  reduction  and  decolorisation  of  a  given 
quantity  of  methylene  blue  by  a  given  volume  of  milk. 

The  test  may  be  carried  out  as  follows  :  40  c.c.  of 
milk  are  introduced  into  a  test  tube  of  about  50  c.c. 
capacity  which  has  been  carefully  cleaned  and  then 
sterilised  by  heating  to  140°  for  about  20  minutes.  I  c.c. 
of  a  0-35  per  cent,  solution  of  pure  methylene  blue  in 
water  is  then  added  and  well  mixed  with  the  milk.  The 
tube  is  immersed  in  a  water  bath  which  is  kept  between 
38°  and  40°,  and  the  time  taken  for  the  blue  colour  to 
entirely  disappear  is  noted.  The  milk  should  not  be 
shaken  while  the  reduction  is  proceeding,  or  the  blue 
colour  may  be  partially  restored  owing  to  atmospheric 
oxidation  ;  this  source  of  error  may  be  guarded  against 
by  covering  the  surface  with  a  thin  layer  of  paraffin  oil. 
If  the  time  taken  for  reduction  is  from  8  to  20  hours,  the 
sample  is  fresh  and  clean,  if  under  5  hours,  the  sample  is 
old  or  unduly  contaminated  with  micro-organisms,  and 
if  under  i  hour,  the  sample  is  either  very  old,  or  it  has 
been  very  carelessly  treated  as  regards  cleanliness. 
This  test  may  be  applied  simultaneously  to  a  number  of 
samples,  the  necessary  appliances,  including  an  auto- 
matic measure  for  the  rnethylene  blue  solution,  tubes, 
stands  for  the  latter  and  water  bath,  being  obtainable 
from  dealers  in  milk  testing  apparatus.  If,  however, 
only  a  few  tests  are  to  be  carried  out,  the  necessary 
appliances  may  easily  be  improvised  from  the  ordinary 
laboratory  equipment.  It  will  be  found  instructive  to 
compare  samples  of  new  and  old  milk  by  this  test. 

Fermenting  Test. — This  test,  which  may  be  carried 
on  concurrently  with  the  reductase  test  (Orla  Jensen), 
has  been  designed  with  a  view  to  ascertaining  the  nature 
of  the  principal  bacterial  contents  of  the  milk.  The  prin- 


220      Industrial  Organic  Analysis 

ciple  involved  is  as  follows  :  If  milk  is  allowed  to  stand 
at  38°  to  40°,  the  development  of  the  micro-organisms 
which  it  contains  is  hastened,  with  the  result  that 
the  struggle  for  existence  between  the  various  types  of 
organisms  at  a  temperature  approximating  that  of  the 
alimentary  canal  will  be  shown  in  a  relatively  short 
time  by  the  appearance  and  smell  of  the  milk.  In 
carrying  out  the  test,  the  milk  is  filled  into  tubes  similar 
to  those  used  in  the  reductase  test,  described  above,  and 
the  tubes  placed  in  a  water  bath  which  is  kept  at  38° 
to  40°.  After  12  hours,  the  samples  are  examined ; 
in  the  case  of  normal,  fresh  milk,  a  sour  taste  and  smell 
will  have  been  developed,  and  in  most  cases  the  coagu- 
lum  will  be  perfectly  even  and  homogeneous,  showing 
that  the  relatively  harmless  lactic  acid  producing 
bacteria  outnumber  the  other  types  of  organisms  which 
may 'occur  in  milk,  and  are  able  to  hold  them  in  check 
for  the  time  being.  If  the  sample  is  unevenly  coagulated, 
showing  separated  whey,  the  lactic  acid  producing 
organisms  have  not  been  able  to  check  the  development 
of  the  protein  hydrolysing  bacteria,  as  will  be  evidenced 
by  a  disagreeable  smell  suggesting  putrefactive  decom- 
position. If  the  proteins  have  been  entirely  dissolved, 
with  the  production  of  a  slightly  turbid  whey,  then  the 
putrefactive  bacteria  will  practically  have  obtained  the 
upper  hand  from  the  start ;  this  may  be  taken  as  a  sign 
that  the  milk  has  certainly  not  been  treated  with  suf- 
ficient care  as  regards  cleanliness  ;  in  this  case  also,  the 
whey  will  have  an  unpleasant  smell,  and  will  probably 
contain  bubbles  of  gas.  A  copious  evolution  of  gas, 
accompanied  by  a  frothy  coagulum  and  a  smell  of 
butyric  acid,  indicates  that  butyric  acid  producing 
organisms  have  obtained  the  upper  hand ;  this  must 
also  be  looked  on  as  a  very  bad  sign. 


Milk  and  Butter  221 

Both  the  reductase  and  fermenting  tests  should  be 
carried  out  with  a  number  of  samples  of  fresh  and  stale 
milk,  in  order  that  the  various  types  of  fermentation 
may  be  observed  ;  the*  same  samples  may  be  used  for 
both  tests,  as  the  addition  of  the  methylene  blue  does 
not  affect  the  results  of  the  fermenting  test.  Further, 
pasteurised  and  unpasteurised  milk  may  be  compared 
by  the  fermenting  test.  From  what  has  been  said  on  the 
subject  of  pasteurising,  it  is  obvious  that  pasteurised 
milk  should  produce  a  more  or  less  flocculent  or  uneven 
coagulum,  and  develop  a  bad  smell. 

CONDENSED  MILK. 

Condensed  milk  is  manufactured  by  evaporating 
fresh  milk  in  vacua  at  temperatures  of  about  40°  to  50° 
to  a  quarter  or  a  third  of  its  original  bulk.  Cane  sugar 
is  often  added  as  a  preservative  to  prevent  the  develop- 
ment of  micro-organisms,  so  that  the  product  may  keep 
indefinitely  in  closed  vessels.  If  cane  sugar  is  not  added, 
it  is  necessary  to  sterilise  the  condensed  milk  after  tin- 
ning. When  condensed  milk  is  diluted  with  3  to  4  times 
its  bulk  of  water,  it  gives  a  product  which  has  the  same 
composition  and  general  properties  as  new  milk,  except- 
ing for  the  presence  of  cane  sugar  which  is  usually  added 
in  sufficient  quantity  to  produce  a  decidedly  sweet  taste. 

ANALYSIS  OF  CONDENSED  MILK. 

Before  taking  the  sample  for  analysis,  the  contents  of 
the  tin  should  be  well  mixed  by  stirring.  After  diluting 
in  accordance  with  the  instructions  on  the  tin,  the 
determination  of  the  various  constituents  may,  in  some 
cases,  be  proceeded  with  as  in  the  case  of  ordinary 
milk,  and  the  results  calculated  back  to  percentages  on 
the  original  condensed  milk  ;  in  the  presence  of  cane  or 
invert  sugar,  special  methods,  to  be  described  below, 


222      Industrial  Organic  Analysis 

must  be  adopted  for  the  estimation  of  the  carbohydrates. 
An  unsugared  product  may  be  diluted  to  a  specific 
gravity  of  1-032  at  15°  in  the  case  of  whole  milk,  or  to 
1-036  in  the  case  of  skim  milk. "  If  cane  or  invert  sugar 
be  present,  the  specific  gravity  can  neither  be  used  as 
an  index  for  diluting  the  condensed  milk  to  a  strength 
corresponding  to  that  of  ordinary  milk,  nor  for  the 
calculation  of  the  solids  less  fat  as  described  above. 

In  the  following  table  will  be  found  typical  analyses 
of  different  varieties  of  condensed  whole  and  skim  milk, 
by  Buttenberg  and  others  (quoted  from  W.  Grimmer)  :— 


Without  Cane  Sugar. 

Condensed  Skim  Milk. 

— 

With  Cane 
Sugar. 

Evapo- 
rated to 

Evapo- 
ratsd  to 

With 
Cane 

Without 
Cane 

4> 

One-Half. 

Third. 

Sugar. 

Sugar. 

Water    . 

27-88 

76-70 

66-91 

27H3 

69-0 

Fat 

9-62 

6-80 

9'75 

0-29 

0-30 

Proteins 

10-27 

5-89 

8-95 

U'59 

12-4 

Lactose 

14-20 

9-13 

12-50 

13-60 

15-7 

Cane  sugar 

36-06 

— 

44-92 

— 

Ash 

1-97 

I-48 

1-89 

2-17 

2-6 

Carbohydrates. — Besides  lactose  and  cane  sugar,  or 
sucrose,  it  is  possible  that  the  condensed  milk  may  con- 
tain invert  sugar,  i.e.,  the  mixture  of  glucose  and  fructose 
obtained  by  the  hydrolysis  of  sucrose. 

The  lactose  is  estimated  in  the  diluted  sample  as 
described  above  for  ordinary  milk.  As  has  already  been 
mentioned,  sucrose  does  not  itself  reduce  Fehling's 
solution,  and  its  presence  will  therefore  not  affect  the 
determination  of  the  lactose  by  the  gravimetric  method. 
Invert  sugar,  on  the  other  hand,  reduces  Fehling's 
solution,  and  will,  if  present,  be  determined  together 
with  the  lactose,  in  which  case  it  will  be  necessary  to 
determine  the  latter  separately  by  another  method 
which  will  be  described  below. 


Milk  and  Butter  223 

The  total  sugar  content  may  be  approximately  esti- 
mated by  subtracting  the  sum  of  the  fat,  proteins  and 
ash,  determined  in  the  diluted  sample,  as  described 
above  for  ordinary  milk,  from '  the  total  solids,  deter- 
mined directly  by  evaporation,  also  as  described  above. 
Subtracting  the  percentage  of  lactose  from  that  of  the 
total  sugar,  the  content  of  sucrose  will  be  found  with 
sufficient  accuracy  for  most  purposes.  If  greater 
accuracy  is  desired,  the  sucrose  must  be  determined 
separately,  as  described  below. 

If  the  amount  of  reducing  sugar  found  would  corre- 
spond to  a  larger  proportion  of  lactose  relatively  to  the 
proteins  and  ash  in  the  sample  than  would  be  expected 
in  ordinary  milk,  then  the  presence  of  invert  sugar  may 
be  suspected.  If,  on  the  other  hand,  the  amount  of 
reducing  sugar  found  corresponds  with  a  normal  amount 
of  lactose  in  proportion  to  the  proteins  and  ash  in  the 
sample,  but  is  appreciably  less  in  amount  than  the  total 
sugar,  then  sucrose  may  be  assumed  to  be  present. 

Qualitative  Test  for  Sucrose. — Quantities  of  not  less 
than  i  per  cent,  of  sucrose  in  condensed  milk  may  be 
detected  by  the  following  method,  due  to  Seliwanoff, 
and  modified  by  Carlsson  :  10  c.c.  of  milk  (or  0-5  gram 
of  lactose  to  be  tested  for  sucrose,  dissolved  in  10  c.c. 
of  water)  are  warmed  in  a  test  tube  with  50  milligrams 
of  resorcin  and  0-5  c.c.  of  25  per  cent,  hydrochloric  acid. 
The  presence  of  sucrose  is  indicated  by  a  red  coloration 
after  boiling  for  a  few  minutes,  no  colour  being  developed 
if  lactose  alone  is  present. 

Determination  of  Sucrose. — The  method  to  be  described, 
which  has  been  elaborated  by  Stokes  and  Bodmer, 
depends  on  the  fact  that  sucrose  may  be  hydrolysed,  or 
inverted,  by  a  dilute  solution  of  citric  acid  while  the  lac- 
tose is  unaffected.  The  principles  involved  have  already 
been  explained  in  Chapter  IV.,  p.  137.  Before  inversion, 
the  reducing  action  on  Fehling's  solution  will  be  due  to 
the  lactose  alone,  or  to  the  lactose  and  invert  sugar,  if 
the  latter  be  present  in  the  original  sample ;  after 


224      Industrial  Organic  Analysis 

inversion,  an  increased  action  on  Fehling's  solution  will 
be  obtained,  owing  to  the  invert  sugar  which  has  been 
produced  from  the  sucrose,  which  itself  is  non-reducing. 
The  increase  in  the  amount  of  copper  oxide  obtained 
after  inversion  may  be  converted  to  sucrose  on  referring 
to  the  table  on  p.  141.  The  reason  for  adopting  the 
following  modified  method  of  inversion  is  that  lactose, 
being  a  disaccharide,  may  be  partially  hydrolysed  to 
galactose,  the  reducing  power  of  which  differs  from  that 
of  the  original  lactose,  if  the  ordinary  method,  involving 
the  use  of  hydrochloric  acid,  be  employed. 

25  c.c.  of  the  diluted  condensed  milk  are  treated  at  the 
ordinary  temperature,  with  sufficient  of  a  I  per  cent, 
solution  of  citric  acid  to  cause  coagulation,  and  made 
up  to  200  c.c.  The  volume  of  the  precipitated  casein, 
including  the  fat,  is  then  calculated  (this  may  be  done 
with  sufficient  accuracy  by  the  method  described  under 
the  *Polarimetric  Determination  of  Lactose  in  ordinary 
Milk,  on  p.  207),  and  an  equal  volume  of  water  added, 
so  that  the  volume  of  the  liquid  itself  is  now  200  c.c. 
The  liquid  is  passed  through  a  dry  filter,  and  the 
reducing  power  of  50  c.c.  is  determined,  as  described  in 
Chapter  IV.,  p.  137.  A  further  50  c.c.  of  the  filtrate 
is  treated  with  0-5  gram  of  citric  acid,  the  latter  being 
dissolved,  boiled  for  fully  30  minutes,  and  the  reducing 
power  determined  as  before.  The  directions  as  to  the 
volume  of  the  solutions,  method  of  heating,  etc.,  given 
on  p.  140,  must  be  strictly  followed  in  both  cases. 

An  alternative  method  for  the  determination  of  sucrose 
in  presence  of  lactose  consists  in  inverting  the  former 
by  means  of  invertase,  an  enzyme  present  in  ordinary 
yeast,  any  action  due  to  the  yeast  itself  or  other  micro- 
organisms being  avoided  by  the  addition  of  a  little 
thymol  as  a  preservative.  In  this  process,  also,  the 
lactose  is  unaffected.  The  method  is  described  fully  in 
Leffman  and  Beam's  "Food  Analysis,"  and  other  works. 


Milk  and  Butter  225 

Determination  of  Lactose  in  Presence  of  Invert  Sugar. — 
As  explained  above,  if  invert  sugar  be  present,  the  lactose 
cannot  be  determined  in  the  ordinary  way,  by  means  of 
Fehling's  solution,  as  the  invert  sugar  also  reduces.  In 
the  following  method,  due  to  Bigelow  and  McElroy,  the 
invert  sugar,  and  also  any  sucrose  which  may  be  present, 
is  fermented  by  yeast  to  alcohol  and  carbon  dioxide  ; 
the  lactose,  which  is  unaffected  by  the  yeast,  may  then 
be  estimated  by  means  of  Fehling's  solution  or  the 
polarimeter. 

250  grams  of  the  condensed  milk  are  diluted  with  water, 
heated  to  boiling,  cooled  to  80°  and  treated  with  a  solu- 
tion of  4  grams  of  glacial  phosphoric  acid  in  a  few 
c.c.  of  water.  After  being  kept  at  80°  for  a  few 
minutes,  the  mixture  is  cooled,  made  up  to  1,000  c.c., 
shaken  and  filtered ;  the  volume  of  the  precipitated 
proteins,  including  the  fat,, is  then  estimated  ;  this  may 
be  arrived  at  with  sufficient  accuracy  by  the  method 
described  under  the  Polarimetric  Determination  of 
Lactose  in  ordinary  Milk,  on  p.  207.  Sodium  hydroxide 
solution  is  added  till  the  solution  is  all  but  neutral,  and 
then  sufficient  water  to  make  up,  together  with  the 
alkaline  solution  just  added,  for  the  volume  of  the  solids. 
For  the  purposes  of  calculation,  the  total  volume  of  the 
liquid  may  then  be  taken  as  1,000  c.c.  After  filtering 
again,  the  solution  is  measured  in  portions  of  100  c.c. 
into  200  c.c.  measuring  flasks.  A  solution  of  20  milli- 
grams of  potassium  fluoride  in  2  or  3  c.c.  of  water  and 
i  gram  of  compressed  brewers'  yeast  is  added  to  each 
flask,  and  the  mixtures  allowed  to  stand  for  10  days  at  a 
temperature  between  25°  and  30°.  The  invert  and  cane 
sugars  will  then  be  fermented  and  destroyed  by  the 
yeast  in  presence  of  the  fluoride,  while  the  milk  sugar 
remains  unaffected.  The  latter  may  then  be  determined 
in  the  solutions,  either  by  the  gravimetric  or  by  the 

I.O.A.  15 


226      Industrial  Organic  Analysis 

optical  method,  as  described  under  the  Determination 
of  Lactose  in  ordinary  Milk.  Before  proceeding  with  the 
actual  determination  by  either  method,  the  contents  of 
the  flasks  should  be  clarified  by  the  addition  of  portions 
of  about  2  c.c.  of  alumina  cream,  prepared  as  described 
below,  made  up  to  200  c.c.,  and  filtered  through  dry 
filters.  For  the  determination  by  Fehling's  solution,  a 
convenient  quantity  of  the  filtrate  may  be  taken  (10  or 
25  c.c.),  so  that  about  150  to  400  milligrams  of  copper,  or 
a  corresponding  amount  of  copper  oxide,  will  be  weighed. 
Alumina  Cream. — This  is  prepared  as  follows  :  A  cold 
saturated  solution  of  alum  is  divided  into  two  unequal 
parts,  and  a  slight  excess  of  ammonia  is  added  to  the 
larger  portion.  The  remainder  of  the  alum  is  added  in 
sufficient  quantity  to  give  a  faintly  acid  reaction. 

Having  determined  the  lactose,  it  is  possible  to  find 
the  amount  of  invert  sugar  present  by  subtracting  the 
lactose  from  the  total  reducing  sugar,  i.e.,  lactose  plus 
invert  sugar,  by  a  simple  calculation  ;  the  amount  of 
copper  oxide  corresponding  to  the  lactose  present  in 
the  sample  taken  for  the  determination  of  the  total 
reducing  sugar  is  calculated  by  referring  to  the  table 
on  p.  205,  and  this  is  subtracted  from  the  actual  amount 
of  copper  oxide  found  in  the  latter  determination.  The 
balance  of  the  copper  oxide  is  due  to  the  reducing  action 
of  the  invert  sugar  alone,  the  amount  of  the  latter  being 
found  by  referring  to  the  table  on  p.  141. 

Calculation  of  the  Composition  of  the  Original  Milk,  and 
the  Degree  of  Concentration  of  the  Condensed  Milk. — The 
fat  contained  in  the  original  milk  before  evaporation 
may  be  approximately  calculated  from  the  non-fatty 
solids  in  the  condensed  milk,  exclusive  of  cane  sugar, 
invert  sugar,  or  other  added  material.  For  the  purposes 
of  the  calculation,  the  percentage  of  non-fatty  solids  in 
ordinary  milk  may  be  assumed  to  be  8-9. 

Then,  the  percentage  of  fat  in  the  original  milk  = 

Fat  in  condensed  milk  x  8*9 
Solids  less  fat  in  condensed  milk 


Milk  and  Butter  227 

Or,  on  the  basis  of  the  protein  content  of  the  condensed 
milk,  assuming  the  protein  content  of  ordinary  milk  to 
be  3-4  per  cent.,  the  percentage  of  fat  in  the  original 
milk  = 

Fat  in  condensed  milk  x  3*4 
Proteins  in  condensed  milk 

In  this  way  it  is  possible  to  determine  whether  the 
condensed  milk  has  been  made  from  whole  milk,  or  milk 
which  has  been  deprived  of  its  fat,  either  wholly  or  in 
part.  The  degree  of  concentration  of  the  condensed 
milk  may  be  arrived  at  by  a  simple  calculation  from  the 
solids  less  fat  or  proteins  contained  therein,  assuming 
the  percentages  of  these  constituents  in  the  original  milk 
to  have  been  8-9  and  3-4  respectively. 

THE  ANALYSIS  OF  BUTTER  AND  MARGARINE. 

The  composition  of  butter  has  already  been  dealt  with 
at  the  beginning  of  this  chapter.  As  a  rule,  the  only 
essential  difference  between  butter  and  margarine  lies 
in  the  composition  of  the  fat.  The  same  methods  as 
are  employed  for  the  determination  of  water,  fat, 
proteins  and  non-fatty  solids  in  butter  may  therefore 
be  applied  to  margarine. 

Sampling. — In  order  to  obtain  a  fair  average  sample  of 
butter  from  a  large  mass,  a  sampling  iron,  such  as  is 
shown  in  Fig.  14,  is  generally  used.  Two  or  three  cylin- 
drical samples  are  taken  in  different  directions  with  the 
iron,  introduced  into  a  wide-mouthed  glass-stoppered 
bottle,  melted  at  a  temperature  not  above  40°,  and 
shaken  vigorously  until  solidified  ;  samples  for  analysis 
may  then  be  taken  from  the  bottle  as  required.  The 
butter  should  not  be  kept  too  long  before  analysis,  as 
it  is  liable  to  alter  in  composition  on  the  development 
of  rancidity. 

Water. — The  estimation  of  water  in  butter  and  mar- 
garine has  already  been  described  in  Chapter  III.,  p.  85. 

15—2 


228      Industrial  Organic  Analysis 


Fat. — The  fat  in  butter  or  margarine  may  be  deter- 
mined by  the  Rose-Gottlieb  method,  as  described  for 
cream,  on  p.  197.  About  2  grams  of  the  sample  are 
weighed  out  on  a  piece  of  fat-free 
paper,  and  introduced  into  the 
graduated  cylinder.  After  adding 
sufficient  hot  water  to  make  up  the 
volume  to  10  c.c.,  and  completely 
melting  the  sample  by  immersing  the 
cylinder  in  hot  water,  i  c.c.  of  the 
ammonia  solution  and  10  c.c.  of  95 
per  cent,  alcohol  are  added.  When 
the  whole  has  cooled  to  the  ordinary 
temperature,  25  c.c.  of  ether  are 
added,  the  contents  are  shaken, 
25  c.c.  of  petroleum  ether  are  added, 
the  contents  shaken  again  and 
allowed  to  stand  until  separated  into 
two  well-defined  layers.  The  pro- 
cedure is  the  same  as  that  described 
above  for  the  determination  of  fat  in 
cream,  the  total  fat  being  removed 
and  weighed  by  extracting  several 
times  with  the  mixture  of  ether  and 
petroleum  ether. 

Instead  of  directly  determining  the 

FIG.  14-Sampling    fat  ^y  the  ab°ve   metl™d   or   other 
Iron.  methods   involving    extraction   with 

organic  solvents,  the  water  and  non- 
fatty  solids,  including  added  salt  or  other  preservative, 
may  be  subtracted  from  100,  the  balance  being  taken 
as  fat. 

In  Germany,  the  minimum  percentage  of  fat  in  sale- 
able butter  is  fixed  by  law  at  80,  and  the  maximum 
water  percentage  at  18  for  unsalted,  and  16  for  salted 


Milk  and  Butter  229 

butter.  In  the  United  Kingdom,  no  definite  minimum 
fat  percentage  has  been  laid  down  by  the  law,  but,  as 
mentioned  above,  the  maximum  water  percentage  is 
fixed  at  16.  (See  Chapter  III.,  p.  85.) 

Non-Fatty  Solids. — These  include  casein,  or  curd,  and 
small  amounts  of  lactose,  lactic  acid  and  inorganic  salts 
derived  from  the  milk  or  cream,  as  well  as  any  added 
salt  or  boric  acid  preservative. 

Barthel,  in  his  "  Methods  of  Analysis  of  Milk  and 
Dairy  Products,"  recommends  the  following  method  for 
their  determination  : — 

5  to  10  grams  of  butter  (or  10  to  15  grams  of  unsalted 
butter)  are  weighed  out  in  a  small  beaker  or  glass  dish 
furnished  with  a  spout,  melted  at  a  gentle  heat,  and 
treated  with  petroleum  ether.1  The  fat  solution  is 
poured  off  through  a  tared  filter,  care  being  taken  that 
all  of  the  aqueous  portion  remains  in  the  beaker  or  dish. 
A  further  quantity  of  petroleum  ether  is  then  added 
and  carefully  decanted  off  through  the  filter,  as  before. 
The  greater  part  of  the  fat  will  now  have  been  removed 
from  the  water  and  non-fatty  solids  ;  it  is  recommended 
that  the  latter  be  not  entirely  freed  from  fat  at  this  stage, 
or  they  will  be  difficult  to  remove  from  the  glass  vessel 
after  drying.  The  non-fatty  solids  are  dried  by  placing 
the  glass  vessel  in  which  they  are  contained  in  an  oven, 
kept  at  100°,  for  2  hours.  They  are  then  completely 
transferred  to  the  tared  filter  by  means  of  petroleum 
ether,  and  washed  with  the  latter  solvent  until  free  from 
fat,  a  drop  of  the  filtrate  leaving  no  grease  spot  when 
evaporated  on  filter  paper.  The  filter  containing  the 
non-fatty  solids  is  then  dried  and  weighed.  Instead  of 
an  ordinary  filter,  a  Gooch  crucible  in  connection  with  a 
filter  pump  may  be  used  with  advantage. 

Proteins. — The  proteins  are  determined  by  means  of 

i  Ether  cannot  be  used  for  this  purpose,  as  it  invariably  dissolves  some  of  the 
salt. 


230      Industrial  Organic  Analysis 

the  Kjeldahl  process,  which  is  described  in  Chapter  I., 
p.  22,  the  amount  of  nitrogen  found  being  multiplied 
by  6-37.  The  determination  cannot  be  conveniently 
carried  out  on  the  butter  itself ;  the  non-fatty  solids 
obtained  from  about  15  grams  of  butter,  as  described 
above,  generally  being  used  for  this  purpose  ;  the  filter 
paper,  which  should  be  free  from  nitrogen,  or  of  known 
nitrogen  content,  containing  the  non-fatty  solids,  may 
be  transferred  direct  to  the  Kjeldahl  digestion  flask,  and 
treated  in  the  usual  way. 

Ash,  Salt  and  Boric  Acid  Preservative. — The  non-fatty 
solids,  obtained  as  described  above,  are  ignited  in  a 
platinum  dish  or  crucible,  with  the  filter,  or  on  the  Gooch 
crucible,  till  all  the  organic  matter  has  been  burnt  off, 
and  the  residue,  free  from  carbon,  is  weighed.  The 
ignited  residue  is  then  dissolved  in  distilled  water,  made 
up  to  a  definite  volume,  filtered,  if  necessary,  through  a 
dry  filter,  and  the  salt  determined  in  an  aliquot  portion 
(about  half  of  the  total)  by  titration  with  decinormal  or 
twentieth  normal  silver  nitrate  solution,  using  potassium 
chromate  as  indicator,  as  is  described  in  most  text-books 
on  quantitative  inorganic  analysis.  The  boric  acid  may 
be  determined  in  butter  or  margarine  by  titration  with 
sodium  hydroxide  solution  in  presence  of  glycerol,  as 
described  in  Chapter  VIII.,  p.  287. 

Subtracting  the  salt  and  boric  acid  (if  any)  from  the 
total  ignited  residue,  the  balance  should  be  the  ash 
derived  from  the  natural  mineral  constituents  of  the 
butter.  If  these  do  not  fall  within  about  0-2  per  cent, 
of  the  butter,  the  latter  may  possibly  contain  other 
added  mineral  matter. 

Methods  for  detecting  other  preservatives  in  butter 
and  margarine  will  be  described  in  Chapter  VIII. 

Examination  of  the  Fat. — The  examination  of  butter 
fat  has  already  been  dealt  with  in  Chapter  III.,  pp.  119 


Milk  and  Butter  231 

et  seq.,  where  the  problem  of  distinguishing  margarine 
fats  from  butter  fat,  and  the  detection  of  the  former  in 
the  latter,  has  received  attention.  As  a  rule,  the  only 
important  distinction  between  butter  and  margarine 
lies  in  the  composition  of  the  fat. 

Pasteurising  Test. — A  test  by  which  it  is  possible  to 
ascertain  whether  the  cream  from  which  the  butter  was 
made  has  been  properly  pasteurised  is  given  on  p.  217. 

Colouring  Matter. — Small  amounts  of  colouring  matter 
are  added  to  margarine,  and  also  to  butter,  in  order  that 
it  may  resemble  the  grass  butter  of  the  summer  months, 
all  the  year  round.  The  vegetable  dyes  turmeric  and 
anatto,  as  well  as  certain  azo-dyes,  are  employed  for 
this  purpose  ;  these  are  harmless,  and  added  in  very 
small  amounts.  Various  methods  by  which  they  may 
be  detected  and  identified,  if  desired,  are  given  in 
the  works  mentioned  below. 


REFERENCES. 

"  Methods  used  in  the  Examination  of  Milk  and  Dairy  Pro- 
ducts," by  C.  Barthel,  translated  by  Goodwin.  1910,  Macmillan 
&  Co.,  London. 

"  Food  Inspection  and  Analysis,"  by  Leach.  1909,  Wiley 
and  Sons,  New  York. 

"  Milk  and  its  Relation  to  Public  Health,"  by  various  authors. 
1908,  Washington  Government  Printing  Office. 

"  Foods,  their  Composition  and  Analysis,"  by  A.  and  M. 
Wynter  Blyth.  1909,  Chas.  Griffin,  London. 

"  Select  Methods  of  Food  Analysis,"  by  Leffmann  and  Beam. 
1905,  Rebman,  Ltd.,  London. 

"  Chemie  und  Physiologic  der  Milch,"  by  W.  Grimmer.  1910, 
Verlagsbuchhandlung,  Paul  Parey,  Berlin. 

"  Mykologie  der  Milch,"  by  Weigmann.  1911,  M.  Hensius 
Nachfolger,  Leipzig. 

"  Milchwirtschaf t, "  by  W.  Kirchner.  1907,  Verl.  von  Paul 
Parey,  Berlin. 

"  Fatty  Foods,"  by  Bolton  and  Revis.  1913,  J.  and  A. 
Churchill,  London. 

"Die  Bacteriologie  in  der  Milch wirtschaft,"  by  Orla  Jensen. 
1913,  Gustav  Fischer,  Jena. 


CHAPTER  VII 

STARCH    AND    ITS    DECOMPOSITION    PRODUCTS — FLOUR, 
BARLEY  AND  MALT 

INTRODUCTORY. 

STARCH  is  one  of  the  most  important  and  character- 
istic products  of  the  vegetable  kingdom,  not  only  on 
account  of  the  prominent  role  which  it  plays  in  plant 
physiology,  but  also  on  account  of  its  great  economic 
value  as  a  food  material  for  animals  and  human  beings. 
A  considerable  portion  of  the  non-nitrogenous  reserve 
food  material  of  plants  is,  in  many  cases,  stored  in  the 
form  of  starch.  The  most  abundant  supplies  of  this 
substance  are  to  be  found  in  seeds,  tubers,  etc.,  where 
it  remains  as  such  until  it  can  be  assimilated  by  the 
embryo  plant.  Starch  is  one  of  the  most  important 
constituents  of  the  flours  used  in  making  bread  and  many 
other  common  articles  of  food,  of  rice,  which  practically 
forms  the  staple  diet  of  one-third  of  the  human  race, 
of  sago,  of  tapioca,  maize  and  barley,  and  of  such 
vegetables  as  potatoes,  beans  and  peas.  As  a  food 
starch  is  used  together  with  the  nitrogenous,  fatty,  saline 
and  other  matter  with  which  it  is  associated  in  nature. 

Starch  which  has  been  separated  from  the  other 
constituents  of  the  grain,  usually  by  washing  with  water 
and  levigation,  is  used,  as  such,  for  laundry  purposes,  as 
a  thickening  material  in  calico  printing,  and  for  the 
preparation  of  adhesive  paste. 

As  regards  indirect  uses,  starch  is  of  importance  as  a 
source  of  alcohol,  which  results  from  the  action  of  yeast 
on  the  sugar  produced  from  the  starch  through  the 
hydrolysing  action  of  mineral  acids,  the  enzyme  diastase, 
or  certain  moulds.  In  the  manufacture  of  beer,  most 
of  the  starch  contained  in  the  malt  is  partially  hydrolysed 


Starch  and  its  Products         233 

by  the  action  of  diastase  (an  enzyme  or  mixture  of 
enzymes  occurring  in  barley  and  other  seeds)  to  dextrin 
and  carbohydrates  of  a  similar  nature,  only  part  of 
the  starch  being  converted  into  sugar  which  can  be 
fermented  to  alcohol  by  the  action  of  yeast.  In  the 
manufacture  of  alcohol  and  spirits,  on  the  other  hand, 
the  starch  is  practically  completely  converted  into  alcohol, 
as  indicated  above.  The  nature  of  the  chemical  changes 
undergone  by  starch  in  the  manufacture  of  beer  and 
spirits  will  be  more  fully  described  when  the  analytical 
processes  applicable  to  starch  and  malt  are  dealt  with. 

The  starches  from  maize,  wheat,  rice  and  the  potato 
are  usually  employed  for  the  direct  applications ;  barley 
is  used  in  the  manufacture  of  beer,  whiskey  and  other 
spirits,  while  dextrin  and  starch  sugar,  for  brewing 
purposes  or  conversion  into  alcohol,  are  generally  pre- 
pared from  starch  from  the  potato  or  from  damaged  or 
inferior  rice,  maize,  etc. 

PROPERTIES  AND  IDENTIFICATION  OF  STARCH. 

Starch  exists  in  plant  cells  in  the  form  of  well-defined 
granules,  varying  in  diameter  from  0-002  to  0-185  mm- 
When  viewed  under  a  microscope  of  fairly  low  magni- 
fying power  (about  150  to  300  diameters)  the  granules 
show  a  very  characteristic  structure,  consisting  of  a 
hilum,  or  nucleus,  surrounded  by  a  number  of  concentric 
rings.  In  order  to  show  up  these  peculiarities  in  struc- 
ture, which  vary  to  a  certain  extent  with  starches  from 
different  sources,  a  small  quantity  of  the  powdered 
starch  should  be  mounted  in  a  dense  medium,  such  as 
glycerine  or  chloral  hydrate,  and  viewed,  if  possible, 
under  oblique  illumination.  With  practice,  it  is  often 
possible  to  determine  the  source  of  a  starch  from  the 
shape  of  the  granules  and  the  relation  of  the  concentric 
rings  to  the  hilum.  Further,  on  examination  under  a 
polarising  microscope  with  crossed  Nicol  prisms,  many 
starches  display  characteristic  arrangements  of  dark 
bands,  usually  giving  the  appearance  of  a  Maltese  cross. 
If  a  selenite  plate  be  placed  between  the  lower  polarising 
Nicol  prism  and  the  object,  colours  will  be  shown  with 
many  starches. 


234      Industrial  Organic  Analysis 

The  microscopic  identification  of  starches  from  different 
sources  will  not  be  exhaustively  treated  of  here  ;  the 
student  wishing  to  pursue  the"  matter  further  should 
consult  Allen's  "Commercial  Organic  Analysis,"  Wynter 
Blyth's  "  Foods,  their  Composition  and  Analysis/'  or 
the  other  works  mentioned  at  the  end  of  this  chapter. 
Starch  in  general  may,  however,  easily  be  identified  by 
a  simple  microscopic  examination,  though  it  should  be 
noted  that  starch  which  has  been  treated  with  water,  or 
exposed  to  a  high  temperature,  as  in  bread  making,  will 
have  lost  its  characteristic  structure,  owing  to  the 
disruption  of  the  granules.  If  necessary,  colouring 
matter  may,  according  to  Allen,  be  removed  in  most 
cases  by  extraction  with  cold  water  and  alcohol. 

Starch  may  readily  be  identified  by  the  characteristic 
blue  colour  which  it  gives  with  iodine.  This  reaction  is 
best  carried  out  by  adding  to  a  solution  of  the  starch  a 
dilute  solution  of  iodine  in  potassium  iodide  solution, 
drop  by  drop  ;  or,  if  the  starch  is  being  observed  under 
the  microscope,  a  drop  of  the  iodine  solution  may  be 
placed  at  the  side  of  the  cover  slip,  so  that  it  will  gradually 
diffuse  into  the  glycerine  in  which  the  granules  are 
mounted,  when  the  latter  will  be  seen  to  take  up  a  dark 
blue  colour. 

Chemically  considered,  starch  is  not  a  definite  sub- 
stance, but  consists  of  a  mixture  of  two  carbohydrates, 
granulose  and  starch  cellulose,  or  rather,  a  series  of 
substances  which  may  be  looked  on  as  gradations  between 
the  two. 

Starch  cellulose  is  a  body  intermediate  as  regards 
complexity  of  structure,  stability  and  other  properties, 
between  granulose  and  ordinary  cellulose  ;  from  the 
latter  it  differs  in  being  convertible  into  soluble  starch 
by  boiling  in  water  or  digesting  in  caustic  alkali  solution. 
It  is  insoluble  in  cold  water  and  saliva,  and  gives  a  yellow 
colour  with  iodine.  Granulose,  on  the  other  hand,  is 
soluble  in  saliva,  and  gives  the  characteristic  blue  colour 
with  iodine.  Ordinary  starch  is  insoluble  in  cold  water  ; 


Starch  and  its  Products         235 

on  heating  with  water,  however,  the  granules  begin  to 
swell  as  the  temperature  approaches  60°,  and  on 
further  heating  they  become  disintegrated  and  mix  with 
the  water,  giving  a  solution  which  gelatinises  strongly 
on  cooling,  and  from  which  starch  may  be  precipitated 
on  the  addition  of  alcohol.  The  starch  thus  obtained 
is  soluble  in  cold  water,  and  is  usually  known  under  the 
name  of  soluble  starch.  The  soluble  variety  is  also 
produced  when  starch  is  heated  in  closed  vessels  to  100°. 
Considered  as  a  member  of  the  carbohydrate  group, 
starch  ranks  after  cellulose  in  point  of  complexity, 
stability  towards  hydrolysing  agents,  and  solubility ; 
when  acted  on  by  ordinary  diastase,  it  is  converted, 
first  into  dextrin,  or  rather  a  series  of  bodies  known 
collectively  as  the  dextrins,  and  ultimately  into  dextrin 
and  the  reducing  disaccharide  maltose.1  Boiling  dilute 
mineral  acids  convert  starch  into  dextrins,  maltose, 
and  finally  into  d  glucose,  or,  as  it  is  commonly  called, 
dextrose.  The  mechanism  of  these  hydrolyses  and  the 
nature  of  the  products  formed  will  be  further  discussed 
when  the  methods  for  estimating  starch  are  explained. 

METHODS    FOR    ESTIMATING    STARCH    IN    COMMERCIAL 
STARCHES,  FLOURS,  POTATOES,  BARLEY,  MALT,  ETC. 

The  following  methods  have  been  chosen  with  a  view 
to  illustrating  the  nature  of  the  decompositions  undergone 
by  starch  under  the  influence  of  various  hydrolysing 
agents.  They  include  (i)  the  hydrochloric  acid  method, 
in  which  the  starch  is  converted  to  glucose,  and  estimated 
as  such  ;  (2)  the  diastase  method,  in  which  the  starch 
is  determined  as  dextrin  and  maltose  ;  (3)  the  com- 
bined diastase  and  yeast  method,  in  which  the  starch  is 
estimated  as  alcohol. 

(i)  Hydrochloric  Acid  Method. — This  method,  which 
is  prescribed  by  the  American  Association  of  Official 
Agricultural  Chemists  (A.O.A.C.),  depends  on  the  fact 
that  starch  is  completely  converted  into  dextrose  on 
boiling  with  dilute  hydrochloric  acid.  The  dextrose 

1  The  Author  is  informed  by  Mr.  W.  A.  Davis  that  if  Taka  diastase  is  used 
dextrose  and  maltose  are  rapidly  formed. 


236      Industrial  Organic  Analysis 

formed  may  be  estimated  by  means  of  Fehling's  solution, 
the  polarimeter,  or  even  roughly  by  the  specific  gravity 
of  its  solution,  and  furnishes  a  measure  of  the  starch 
present  in  the  sample.  It  should  be  noted  that  a  certain 
class  of  gummy  bodies  known  as  the  pentosans,  which 
are  often  present  in  notable  proportions  in  flours,  give 
rise  to  reducing  sugars,  i.e.,  the  pentoses  (sugars  contain- 
ing 5  carbon  atoms  in  the  molecule),  on  boiling  with 
acids.  The  method  is  therefore  only  applicable  to  com- 
mercial starches  which  have  been  freed  from  extraneous 
gummy  matter,  such  as,  for  example,  the  starches  used 
for  stiffening  textiles. 

3  grams  of  the  sample  are  mixed  with  about  50  c.c.  of 
cold  water,  stirring  at  intervals  for  about  an  hour.  The 
insoluble  residue  is  then  collected  on  a  filter  and  washed 
with  water  until  the  filtrate  measures  250  c.c.  The 
residue,  which  has  now  been  freed  from  soluble  carbo- 
hydrajtes,  is  heated  on  a  water  bath  for  2j  hours  with  a 
mixture  of  200  c.c.  of  water  and  20  c.c.  of  hydrochloric 
acid  of  specific  gravity  1-125  (i-e->  2'5  Per  cent.  HC1),  in 
a  flask  fitted  with  a  reflux  condenser.  The  mixture  is  then 
cooled,  neutralised  with  sodium  carbonate,  made  up  to 
250  c.c.,  and  filtered  through  a  dry  filter.  The  dextrose 
is  determined  in  an  aliquot  portion  of  the  filtrate  by  means 
of  Fehling's  solution,  as  described  in  Chapter  IV.,  p.  141. 
The  weight  of  dextrose  found,  multiplied  by  0-9,  gives 
the  weight  of  the  starch.  This  method  gives  results 
3  to  5  per  cent,  too  low,  owing  to  destruction  of  some  of 
the  dextrose  by  the  hydrochloric  acid. 

(2)  Diastase  Method. — In  this  method  the  starch  is 
freed,  as  far  as  possible,  from  extraneous  matter,  and 
hydrolysed  by  means  of  the  enzyme  diastase  to  dextrin 
and  maltose,  both  of  which  are  then  estimated.  The 
advantage  over  the  foregoing  method  lies  in  the  circum- 
stance that  the  pentosans  and  other  gum-like  products 
which  are  converted  into  reducing  sugars  by  hydro- 


Starch  and  its  Products         237 

chloric  acid  are  not  affected  by  diastase  ;  it  may  be 
applied  to  the  estimation  of  starch  in  flours,  grain, 
potatoes,  malt,  etc. 

Before  giving  the  working  details,  a  few  words  must 
be  said  regarding  the  rationale  of  the  process.  The 
hydrolysing  action  of  ordinary  diastase  on  starch  pro- 
ceeds with  fair  rapidity  until  four-fifths  of  the  starch 
molecule  have  been  converted  into  maltose,  and  one-fifth 
into  dextrin,  the  resulting  mixture  being  composed  of 
80-8  per  cent,  of  maltose  and  19-2  per  cent,  of  dextrin. 
After  this  stage  has  been  reached  the  action  slows  down 
to  such  an  extent  that  it  is  advantageous  to  stop  it 
altogether,  and  proceed  with  the  determination  of  the 
maltose  and  dextrin,  which  will  give  a  measure  of  the 
amount  of  starch  originally  present. 

The  changes  which  take  place  may  be  represented 
by  the  following  equations  (A.  J.  Brown's  "  Laboratory 
Studies  for  Brewing  Students  ")  : — 

1.  5  (C12H20010)  +  3  H20  =  2  C12H22On  + 

Soluble  starch.  Free  maltose.     Malto-dextrin. 

-j-  CiaH^Ojo. 
Stable  dextrin. 

2.  (r^o")    +  H20  =  2  CM&n. 

VCxa-H^Mo/ 

Malto-dextrin.  Free  maltose. 

The  final  result  of  the  conversion  may  thus  be  repre- 
sented by — 

4  CiaH»Ou     -[-    C12H2oO10. 

Maltose.  Stable  dextrin. 

It  will  be  noticed  that  the  malto-dextrin  molecule, 
represented  as  being  built  up  of  maltose  and  dextrin 
molecules,  is  completely  converted  into  maltose,  whereas 
free  dextrin,  here  referred  to  as  stable  dextrin  to  dis- 
tinguish it  from  malto-dextrin,  is  unaffected  under  the 
conditions  of  the  experiment. 

Regarding  the  products  of  the  conversion,  maltose, 
being  a  typical  reducing  sugar,  may  be  estimated  by 
means  of  Fehling's  solution.  Dextrin,  wjien  pure,  is  an 


238       Industrial  Organic  Analysis 

amorphous  white  solid,  soluble  in  cold  water  and 
possessing  an  optical  activity  of  [a]D  =  202-0°.  It  has  no 
action  on  Fehling's  solution,  so  that  an  estimation  of  the 
reducing  power  of  the  product  of  the  starch  conversion, 
together  with  its  optical  activity,  will  give  the  amounts 
of  maltose  and  dextrin  present  by  a  simple  calculation. 
Maltose  may  be  separated  from  dextrin  by  treatment 
with  alcohol,  in  which  the  former  is  soluble  and  the 
latter  insoluble. 

The  method  described  is  due  to  C.  O'Sullivan  (J.C.S., 
1884,  p.  2).  Before  dealing  with  the  actual  determina- 
tion, two  preliminary  operations  must  be  described  : 
(a)  preparation  of  diastase  or  malt  extract,  and  (b)  pre- 
paration of  the  sample. 

(a)  Preparation  of  Diastase  or  Malt  Extract. — Either 
diastase  or  malt  extract,  prepared  as  directed  below,  may 
be  used  for  effecting  the  hydrolysis  of  the  starch. 

(i.)  Diastase. — 2  to  3  kilos,  (or  a  smaller  quantity  if 
desired)  of  finely  ground  barley  malt  are  steeped  in 
water,  just  sufficient  to  cover  the  whole.  After  standing 
for  several  hours,  the  mash  is  filtered  by  pressing 
through  calico,  and  the  filtrate,  if  turbid,  is  passed 
through  an  ordinary  filter.  The  diastase  is  now  pre- 
cipitated from  the  clear  filtrate  by  adding  alcohol  of 
specific  gravity  0-83,  so  long  as  the  precipitate  produced 
is  flocculent,  the  addition  being  discontinued  as  soon  as 
the  liquid  begins  to  appear  milky  or  opalescent.  The 
diastase  is  filtered  off  and  washed  with  alcohol  of 
specific  gravity  0-86  to  0-88,  dehydrated  by  further 
washing  with  a  little  absolute  alcohol,  and  dried  over 
sulphuric  acid  in  a  desiccator.  Diastase  thus  prepared 
is  a  friable,  white  soluble  powder  which  retains  its 
activity  as  a  hydrolysing  agent  for  starch,  for  a  con- 
siderable time. 

Diastase  is  one  of  the  enzymes  which  convert  the 
starch  present  in  the  seed  into  soluble  carbohydrates 


Starch  and  its  Products         239 

which  may  be  assimilated  by  the  embryo  plant  during 
germination  ;  the  part  which  it  plays  in  brewing,  in 
converting  starch  into  maltose  and  products  belonging  to 
the  dextrin  group,  will  be  referred  to  later. 

(ii.)  Malt  Extract. — 100  grams  of  pale  malt  are 
digested  with  250  c.c.  of  water  at  65°,  and  allowed  to 
stand  at  the  room  temperature  for  12  hours.  After 
filtering,  the  liquid  is  ready  for  use. 

It  is  advisable  to  determine  the  diastatic  capacity  of 
the  diastase  or  malt  extract  used  in  the  starch  deter- 
mination, by  the  methods  given  below  (see  p.  249),  in 
order  to  make  sure  that  a  preparation  of  sufficient 
diastatic  activity  is  being  used. 

(b)  Preparation  of  the  Sample. — The  material,  repre- 
senting a  fair  average  sample,  must  be  freed  from 
extraneous  matter  by  treating  it  with  various  solvents 
which  will  not  act  on  the  starch. 

5  grams  of  the  flour,  or  finely  ground  malt  or  barley, 
are  soaked  with  alcohol  of  specific  gravity  0-82,  and  20  to 
30  c.c.  of  ether  are  added.  The  flask  containing  the  mix- 
ture is  corked  and  set  aside  for  a  few  hours,  after  which 
the  whole  is  filtered,  and  the  residue  washed  with  ether. 
This  treatment  will  remove  fatty  matter  from  the  sample. 

The  next  step  is  to  remove  matter  soluble  in  dilute 
alcohol,  which  includes  sugars,  albuminoids  other  than 
casein,  and  other  substances. 

The  fat-free  flour  is  transferred  to  a  flask  and  treated 
with  80  to  90  c.c.  of  alcohol  of  specific  gravity  0-90  at 
35°  to  38°  for  a  few  hours,  shaking  occasionally.  The 
alcoholic  solution  is  passed  through  the  same  filter  as 
was  used  for  the  alcohol-ether  operation  just  described, 
and  the  insoluble  residue  is  washed  on  the  filter  with 
alcohol  of  the  same  strength  and  at  the  same  temperature 
as  already  used  in  the  present  operation. 


240      Industrial  Organic  Analysis 

Finally,  the  water  soluble  material  must  be  removed  ; 
this  includes,  among  other  substances,  the  amylans,  a 
group  of  carbohydrates  having  the  same  empirical 
formula  as  starch,  but  differing  from  the  latter  in  being 
soluble  in  cold  water,  and  giving  no  coloration  with 
iodine.  Wheat,  for  example,  generally  contains  about 
2  to  2 '5  per  cent,  of  /?  amylan. 

The  residue  from  the  last  operation  is  digested  with 
500  c.c.  of  water  at  the  ordinary  temperature  for  24 
hours,  and  then  decanted  through  the  filter  used  pre- 
viously ;  the  insoluble  residue  is  transferred  to  the 
filter  by  means  of  water,  and  repeatedly  washed  with 
water  at  35°  to  38° ;  the  latter  process  may  sometimes 
be  very  tedious. 

(c)  Conversion  of  the  Starch. — The  insoluble  residue 
from  the  last  operation  is  transferred  to  a  beaker,  and 
boiled  with  40  to  45  c.c.  of  water  for  a  few  minutes,  with 
constant  stirring,  in  order  to  convert  the  starch  into  the 
soluble  form.  The  mixture  is  then  cooled  to  62°  to  63°, 
and  0-2  to  0-4  gram  of  active  diastase  dissolved  in  a  few 
c.c.  of  water,  or  10  c.c.  of  an  active  malt  extract, 
prepared  as  described  above,  are  added. 

If  malt  extract  is  used  a  certain  amount  of  sugar  and 
dextrin  will  be  introduced  ;  this  must  be  determined 
by  making  a  blank  experiment  with  an  equal  quantity 
of  the  same  malt  extract,  treating  this  in  exactly  the 
same  manner  as  the  starch  solution,  i.e.,  diluting  to  the 
same  volume  and  keeping  it  at  the  same  temperature 
for  the  same  length  of  time.  The  amount  of  sugar  and 
dextrin  due  to  the  malt  extract  added  to  the  starch  may 
then  be  estimated  and  deducted. 

After  the  addition  of  the  diastase  solution  or  malt 
extract,  the  mixture  is  kept  at  62°  to  63°  in  a  water  bath  ; 
in  a  short  time,  it  will  be  found  that  the  characteristic 


Starch  and  its  Products         241 

blue  coloration  will  no  longer  be  obtained  when  a  drop 
of  the  mixture  is  treated  with  a  drop  of  dilute  iodine 
solution  ;  at  this  point,  all  the  starch  will  have  been 
converted  into  maltose  and  dextrin  ;  it  is,  however, 
recommended  to  continue  the  digestion  for  another  hour, 
in  order  that  the  subsequent  nitration  may  be  facilitated. 
The  contents  of  the  beaker  are  now  boiled  for  8  to  10 
minutes  in  order  to  render  the  diastase  inactive,1  trans- 
ferred to  a  filter,  and  the  filtrate  received  in  a  100  c.c. 
measuring  flask  ;  the  residue  is  carefully  washed  with 
small  quantities  of  water  at  a  time,  the  washings  also 
being  received  in  the  measuring  flask  ;  when  the  latter 
is  nearly  full,  the  contents  are  cooled  to  15-5°  and  made 
up  to  100  c.c.  with  water  at  that  temperature.  Should 
the  volume  of  the  filtrate  exceed  100  c.c.,  it  may  be 
concentrated  by  evaporation.  The  specific  gravity, 
optical  activity  and  reducing  power  of  the  solution  are 
now  determined  in  order  to  find  the  amounts  of  maltose 
and  dextrin  present. 

(d)  Specific  Gravity. — The  object  of  this  determination 
is  to  find  the  amount  of  solid  matter  contained  in  the 
solution  ;  this  should,  if  the  above  processes  have  been 
properly  carried  out,  be  found  to  be  very  nearly  equal 
to  the  sum  of  the  maltose  and  dextrin  as  determined 
by  the  optical  activity  and  reducing  power  of  the 
solution,  together  with  the  added  diastase.  The  specific 
gravity  method  is  very  commonly  employed  in  deter- 
mining the  amount  of  solid  matter  contained  in  solutions 
of  carbohydrates,  in  preference  to  the  method  of  evapo- 
rating and  weighing  the  residue,  owing  to  the  manner  in 
which  these  bodies  retain  water.  The  specific  gravity 
is  determined  in  the  usual  way  by  means  of  a  picnometer 


1  Compare  the  destruction  of  the  hydrogen  peroxide  decomposing  enzyme  in 
milk  by  heating  to  80°  (see  p.  217).    ' 

I.O.A.  16 


242      Industrial  Organic  Analysis 

or  specific  gravity  bottle,. and  for  the  purposes  of  the 
calculation,  expressed  as  the  weight  of  1,000  c.c.  of  the 
solution  in  grams.  The  difference  between  the  specific 
gravity  thus  expressed,  and  1,000,  divided  by  the  number 
of  grams  per  100  c.c.,  gives  what  is  known  as  the  solution 
factor  for  the  given  material. 

Thus      Sp.  gr.  -  1,000    =  solution  facton 
Gms.  per  100  c.c. 

Knowing  the  solution  factor,  the  concentration  of  the 
solution  may  easily  be  calculated  from  the  specific 
gravity.  The  solution  factors  of  the  various  carbo- 
hydrates vary  slightly  with  the  concentrations  of  their 
solutions.  They  have  been  determined  for  solutions  of 
various  carbohydrates  of  varying  strengths  by  Brown, 
Morris  and  Millar  (/.  C.  S.,  1897,  Vol.  Ixxi.,  p.  72). 

For  the  purpose  of  the  present  estimation,  the  solution 
factor"  to  be  used  is  3-95. 

(e)  Determination  of  the  Maltose. — This  is  carried  out 
according  to  the  method  of  Brown,  Morris  and  Millar, 
as  described  on  p.  137. 

As  maltose  is  a  reducing  sugar,  hydrolysis  is  not 
necessary  as  in  the  case  of  cane  sugar.  If  the  directions 
are  adhered  to,  the  amount  of  maltose  present  may  be 
calculated  from  the  amount  of  copper  oxide  formed,  by 
referring  to  the  accompanying  table  on  p.  243.  Dextrin 
has  no  effect  on  Fehling's  solution,  so  that  the  total 
copper  oxide  weighed  will  be  due  to  the  action  of  maltose. 

(/)  Optical  Activity  of  the  Solution. — The  optical 
activity  of  the  solution  of  the  starch  conversion  product 
is  determined  in  the  usual  manner,  the  concentration 
having  been  determined  by  the  specific  gravity. 
At  the  concentrations  dealt  with,  the  optical  activity 
of  dextrin  is  [a]D  =  +  200-4°  and  that  of  maltose 
WD  =  +  138°.  Knowing  the  amount  of  maltose 


Starch  and  its  Products         243 

TABLE  FOR  CALCULATING  THE  AMOUNTS  OF  MALTOSE 

CORRESPONDING  TO  VARYING  AMOUNTS  OF  COPPER. 

(Method  of  Brown,  Morris  and  Millar.) 


Maltose, 
grams. 

Cu, 
grams. 

Maltose, 
grams. 

Cu, 
grams. 

•070 

•0772 

% 

•185 

•2017 

•075 

•0826 

•190 

•2072 

•080 

•0880 

•195 

•2126 

•085 

•0934 

•2OO    ' 

•2l8o 

•090 

•0988 

•205 

•2234 

•095 

•1042 

•210 

•2288 

•IOO 

•1097 

•215 

•2342 

•105 

•II5I 

•22O 

•2397 

•no 

•I2O5 

•225 

•2451 

•ii5 

•1259 

•230 

•2505 

•I2O 

•1313 

•235 

•2559 

•125 

•1367 

•240 

•2613 

•130 

•1422 

•245 

•2667 

•135 

•1476 

•250 

•2722 

•140 

•1530 

•255 

•2776 

•145 

•1584 

•260 

•2830 

•150 

•1634 

•265 

•2884 

•155 

•1692 

•270 

•2938 

•160 

•1747 

•275 

•2992 

•165 

•1801 

•280 

•3047 

•170 

•1855 

•285 

•3101 

•175 

•1909 

•290 

•3155 

•180 

•1963 

•295 

•3209 

•300 

•3264 

•305 

•3318 

Factor  for  converting  CuO   to   Cu  =  0-7989. 

present  from  the  reducing  power  of  the  solution,  the 
optical  activity  due  to  this  constituent  may  be  calculated 

1 6— 2 


244      Industrial  Organic  Analysis 

and  deducted  from  the  total  optical  activity,  the  differ- 
ence being  due  to  the  dextrin. 
The  total  starch  is  equal  to  :— 

Dextrin  +  Maltose  X  0-95. 

As  mentioned  above,  the  maltose  and  dextrin  due  to 
added  malt  extract  must  be  determined  in  a  blank 
experiment  and  allowed  for. 

(3)  Determination  of  Starch  as  Alcohol.  —  Although 
the  stable  dextrin  which  is  formed  as  one  of  the  products 
of  the  action  of  diastase  on  starch  in  the  process  just 
described  is  only  acted  on  by  diastase  with  extreme 
slowness,  and  is  not  fermented  by  yeast  alone,  yet  by 
the  joint  action  of  diastase  and  yeast  it  is  readily  con- 
verted into  alcohol.  Maltose,  on  the  other  hand,  is 
fermented  to  alcohol  by  yeast  in  the  absence  of  diastase. 
In  the  manufacture  of  alcohol,  spirits  or  vinegar,  where 
the  object  is  to  obtain  a  maximum  yield  of  alcohol  from 
the  starch,  yeast  is  allowed  to  act  on  the  product  of  the 
diastatic  conversion  without  destroying  the  diastase 
In  brewing,  on  the  other  hand,  where  the  object  is  to 
prepare  a  product  containing  dextrin-like  products  and 
other  carbohydrates,  with  only  a  relatively  small  pro- 
portion of  alcohol,  the  diastatic  fermentation  is  arranged 
so  that  only  a  small  proportion  of  maltose  is  produced, 
after  which  the  diastase  is  destroyed  by  heating,  and  the 
product  treated  with  yeast ;  in  this  way  the  maltose 
only  is  converted  into  alcohol. 

In  the  analytical  method  under  consideration,  which 
has  been  devised  for  the  estimation  of  starch  in  barley, 
5  grams  of  the  finely  ground  sample  are  extracted,  first 
with  alcohol  and  ether,  and  then  with  alcohol  of  specific 
gravity  0-90,  as  described  above  under  the  Preparation 
of  the  Sample  in  the  Diastase  Method  for  estimating 
Starch.  The  residue  is  washed  into  a  flask  with  water, 
and  thoroughly  boiled  in  order  to  remove  all  traces  of 
alcohol  and  to  gelatinise  the  starch.  The  diastatic 


Starch  and  its  Products         245 

conversion  is  effected  by  means  of  diastase  or  malt 
extract  at  62°,  as  described  above ;  then,  without 
destroying  the  diastase  by  boiling,  the  conversion  pro- 
duct is  cooled  to  26°  to  27°,  and  submitted  to  the 
action  of  i  gram  of  brewers'  yeast  in  a  flask  at  that 
temperature  for  several  days.  The  alcohol  may  then  be 
estimated  in  the  product  by  the  following  method,  which 
is  very  commonly  employed  for  the  estimation  of  alcohol 
in  the  presence  of  non- volatile  matter : — 

The  carbon  dioxide  is  first  removed  by  repeatedly 
pouring  the  liquid  from  one  vessel  to  another,  or  by 
filtration,  and  the  whole  is  made  up  to  100  c.c.,  or,  if 
necessary,  to  a  larger  definite  volume,  with  distilled 
water.  100  c.c.  are  then  distilled  from  a  flask  of  con- 
venient size  connected  with  a  straight  tube  condenser, 
until  80  c.c.  of  distillate  have  been  collected.  The 
distillate  is  made  up  to  100  c.c.  with  distilled  water,  and 
its  specific  gravity  determined  at  60°  F.  by  means  of 
the  picnometer  or  specific  gravity  bottle.1  Reference 
to  the  accompanying  tables  will  show  the  percentage  of 
alcohol  in  the  distillate.  In  order  to  calculate  the 
percentage  of  absolute  alcohol  in  the  sample  before 
distillation,  it  will  be  necessary  to  know  either  its  specific 
gravity  and  volume  or  its  weight,  and  likewise,  also,  the 
specific  gravity  and  volume  or  the  weight  of  the  distillate. 

Then 

Sp.  gr.  of  distillate  x  vol.  of  dist.  in  c.c.   x  %  of  alcohol  in  dist. 
Sp.  gr.  of  sample  x  vol.  of  sample  in  c.c. 

=  Percentage  of  absolute  alcohol  by  weight  contained 
in  the  sample,  or 

Weight  of  dist.   x  %  of  alcohol  in  dist. 
Weight  of  sample  taken. 

1  For  a  description  of  the  exact  determination  of  specific  gravities  of  liquids,  see 
Allen's  "Commercial  Analysis/'  Vol.  I.,  pp.  5  &  6,  1909  edition. 


246     Industrial  Organic  Analysis 


N   N    CO  CO  ^  ON  O    N    CO  rt-  rJ-O  00   M    co  ^OO    COO    O    CO  O  O 

9  T1"0?  7*  ^?  M  ^9  ?  T^"0?  Y"5  9^  7"300  <^MT>9  ?*•  9^  M 

°'S  OOOl^^V^^^^NNo")ro^-MMNN<N 

j£  •< 

"3  "3       ttj  OOOOOOOOOOOOOOOOOOQOOOOOO 

0*0  >>^ 

n     i  -0  -5  ^OO  00    ON  ON  t^OO  00    ON  ON  C^OO  00          ON  t^-00  00    ON  ON  r^OO  00    ON  ON 

MMMMMNNNNNcocOCOcOCO-'4-Tj-T}--«3--<t- 

ON  t^  t^  t^  ON  ONOO  CO  00  OO  00  N  lOOO  O  O  t>  COOO  NO  ^  O  O  O 
lOOifOC^-M  »OO  "~>O  >OO<  t^-M  IOOOOO  M  ro\O  M  C4  CO  ro  -<f 

QNOO  oo  t^  t^-oo  oo  r^  t>-o  oo  r^  o-\c  o  «o  rf  ^  co  n  O^oo  r~^o  »o 

OOOOOOOOOOC^r^l>.t^  C^O  OOOO    'OiO'O'O'OCOrOCOCOfO 

QN  ^  Q\  (^  Qs  ON  ON  0s*  ON  ON  ON  ONONONONONONONONONONONONO^ON 

6666006666666666666666666 

OOOMMrori-iO  t^OO    m  iO  r--  ON  M  O    ON  OOO    O    ^-  O    >O  M    t^ 
9    T1"0?    t*  ^    M    V^  9^  7^  V^  T*"0?    ^  ^   M    ON  rocp    N    r^OO    ro  r^  (N  O 
NMrOro66MMO»cboNON66    t^-00  00    ON  ON 
HMHMNC><(NC^r<NNNrorocorOCOfOrO 

"c"o       aj      I        OOOOOOOOOOOOOOOOOOOOOOOOO 

^•gxS    j 
n£-° 3,       ' 

^•<      ,°      I 

,.  HiOON«OMTt-rOMOONr^-^-Mt^o-NOMrnrON    Tt-O  00  00  00 

.3  ^ .  •  00    M    Tl-OO    (SMOMOOON^J-ON  <T)00    M    Tf  t^  O    rOO   t^OO    ON  O 

NNMOOMOOONONOOON  ONOO  ONOO  Jx  t^O  ro  N  M  O  O 
ONONONONONOOOOCOt^t^r^t^OO'O'O'O'O'OiO'^'^''^-'^'^- 
ON  ON  ONONONONONONONONONONONONONONONONO^ONONONONONON 

6666666666666666666666666 

C^  ON  H    O    o*)O   ON  01  00    fOOO    CO  ON 

's?  9  "^^9  T1"?^ 
t^ob  ob  «oo  o  o 

MHMMMMI-tMMOIC^NC^tNNCOf^NCOCO 

OOOOOOOOOOOOOOOOOOOOOOOOO 

ONt^c^t^-ONMO<r>o^f^io^j'fOMONioNoof>o  t^o  M-  o  o  o 
N  >ooo  H  Tj-oo  ^  t^MO  100  >oo  ^t-ror^o  rj-  r^-oo  O  <N  m  10 
^o  ^  ^  ^  ro  co  ^o  c^  01  M  ro  co  c^c^tHcsuMOON  t^  c*^o  *o  ^ 

ON  ON  ON  ON  ONOO  OOCOOOOO    t^I>-C^t^  t>»O  OOO    lO^-^-Tt-rt--^- 

6666666666666666666666666 

OOONONON^  >OO     O-  t>-O     t^ONM     MOGO     M     T}-t^lOON    Tt-OO     CO 

9    7^  ^*  M    f>  9    7*"°°    NO    NO    9    lOONO    9 

66oHMCoobob6NONOO^^^V^ 

MI-HMMMNN 

OOOOOOOOOOOOOOOOOOOOOOOOO 

>1l5l>^i>>>ir)9 

HIN 

rhrt" 

M  O  COO  M  ITNOO  00  (^  t^O  H  H  M  O  00  »O  t^OO  00  00 
oN"3-r^OOOTi-O^cOOinOlOO  ^-OO  M  O  ONOO  O  IN  ^-O 
O  ONOO  t^  t^O  O  >O  Tf-  Tj-O  iO»OTi-^u~>M  •*•  ro  N  M  M  O  ONOO 
O  ONONONONOOOOOOOOOO  l^t^t^t^f^OOOOO  'OiOiOTj-Tj- 
Q,ONONONONONONONONC?>ONONONONONONONONONONONONONONON 

Mo6666666b666666666666666 


Starch  and  its  Products         247 

=  Percentage  of  absolute  alcohol  by  weight  contained 
in  the  sample.  100  parts  of  alcohol  correspond  to  169-2 
parts  of  starch. 

If  malt  extract  is  useofeto  effect  the  diastatic  conversion 
of  the  starch,  it  will  be  necessary  to  conduct  a  blank 
experiment  with  an  equal  quantity  of  the  extract,  yeast 
and  water,  and  deduct  the  alcohol  formed  from  this 
from  the  total  alcohol  formed  in  the  determination 
proper. 

The  above  method  for  estimating  alcohol  is  commonly 
applied  to  beers,  wines  and  spirits.  It  is  used  largely 
for  the  determination  of  the  "  original  gravity "  of 
fermented  worts,  i.e.,  the  specific  gravity  of  the  wort 
before  fermentation,  for  excise  purposes.  The  "  original 
gravity  apparatus  "  consists  essentially  of  a  distilling 
flask  connected  to  a  vertically  placed  spiral  condenser, 
of  certain  dimensions ;  by  determining  the  specific 
gravities  of  the  still  residue  and  the  distillate,  both 
diluted  to  the  original  volume  of  the  sample,  it  is  possible 
to  calculate  the  number  of  "  degrees  of  specific  gravity  '• 
lost  by  the  wort  during  fermentation. 

In  determining  the  alcohol  in  beer,  100  c.c.  of  the 
sample  may  be  distilled  till  80  c.c.  of  distillate  have  been 
collected.  Spirits  containing  about  50  per  cent,  or  more 
of  alcohol  should  be  diluted  before  distilling  ;  thus, 
50  c.c.  of  the  sample  may  be  diluted  with  100  c.c.  of 
distilled  water,  and  distilled  until  the  distillate  measures 
100  c.c. 

In  order  to  illustrate  the  difference  in  the  amounts  of 
alcohol  formed  from  starch  in  the  case  when  diastase 
and  yeast  are  allowed  to  act  together,  and  in  the  case 
when  the  yeast  is  only  allowed  to  act  after  the  diastase 
has  been  destroyed,  the  above  determination  of  starch 
as  alcohol  may  be  repeated  with  an  equal  quantity  of 


248      Industrial  Organic  Analysis 

material  from  the  same  sample,  with  the  sole  difference 
that  the  mixture  is  boiled  for  10  minutes  in  order  to 
destroy  the  diastase  after  the  diastatic  fermentation  is 
complete  ;  the  yeast  is  to  be  «added  after  cooling  to 
26°  to  27°,  and  allowed  to  act  at  this  temperature  as 
described  above.  All  other  conditions  being  equal,  the 
amount  of  alcohol  formed  in  this  case  should  be  less  than 
in  the  preceding  case. 

MALT  AND  MALT  EXTRACT. 

Under  this  heading,  a  method  is  described  by  which 
the  relative  diastatic  activity  of  diastase  or  malt  extract 
will  be  determined  in  terms  of  what  is  known  as  the 
"  Lintner  Value."  The  value  of  a  malt  extract  for 
brewing  purposes  depends  very  largely  on  its  diastatic 
activity,  i.e.,  on  the  ability  of  the  diastase  which  it 
contains,  to  effect  a  rapid  conversion  of  the  starch  into 
maltose  and  dextrin  ;  for  the  purpose  of  the  last  two 
analytical  processes  described  for  the  determination  of 
starch,  it  is  also  desirable  that  the  diastase  or  malt 
extract  used  should  be  fairly  active.  If  commercial 
malt  extracts  are  employed,  it  is  desirable  that  their 
diastatic  activity  should  be  determined  before  use,  as 
many  preparations  on  the  market  have  little  or  no  power 
to  effect  the  hydrolysis  of  starch. 

Before  the  determination  is  described,  a  few  words 
may  be  said  regarding  the  preparation  and  composition 
of  malt.  Malt  is  prepared  by  steeping  barley  in  water 
and  then  allowing  it  to  germinate  at  a  suitable  tempera- 
ture ;  during  this  process,  diastase,  as  well  as  other 
enzymes,  are  formed  in  order  that  the  starch  may  be 
converted  into  soluble  products  which  can  be  assimilated 
by  the  growing  embryo.  Before  germination  has  pro- 
ceeded to  any  extent,  the  vital  processes  are  stopped  by 
drying  and  curing  the  grain  in  a  kiln  ;  the  action  of  the 
diastase  on  the  starch  and  other  carbohydrates  is  thereby 
stopped  for  the  time  being,  to  be  continued  when  the 
malt  is  subsequently  made  into  a  mash  with  water. 


Starch  and  its  Products 


249 


During  the  drying  process,  various  substances  are  formed 
which  impart  colour  and  flavour  to  the  malt  and  to  the 
products  prepared  from  it. 

Malt  varies  in  colour  from  light  yellow  to  dark  brown, 
according  to  the  origin  of  the  barley  and  the  degree  of 
the  curing,  or  drying.  It  should  be  crisp  and  white 
inside  ;  its  diastatic  activity  depends  very  largely  on 
the  freshness  and  quality  of  the  grain  from  which  it  is 
prepared,  as  well  as  on  the  degree  of  curing. 

The  following  typical  analysis,  from  Blount  and 
Bloxam's  "  Chemistry  for  Engineers  and  Manufac- 
turers," 1  will  give  an  idea  of  the  approximate  com- 
position of  barley  and  malt ;  the  most  noticeable 
differences,  on  the  analytical  figures,  between  barley  and 
malt  are  the  lower  proportion  of  water  and  the  higher 
proportion  of  soluble  carbohydrates  contained  in  the 
latter.  Further,  malt  usually  contains  slightly  less  fatty 
matter,  and  more  soluble  nitrogen  compounds  than 
barley. 


— 

Barley. 

Malt. 

Per  cent. 

Per  cent. 

Water 

14-1 

5'8 

Proteins    . 

10-6 

13-1 

Fat  . 

2-1 

17 

Carbohydrates  . 

637 

65-4 

(mainly  starch). 

(about  J  being  fer- 

mentable sugar) 

Fibre 

7-1 

n-6 

Ash 

2-6 

2-6 

Determination  of  the  Diastatic  Activity  (Lintner  Value) 
of  Malt? — This  determination  is  based  on  Kjeldahl's 
observation  that  in  malt  diastase  conversions,  where  the 


1  See  also  the  figures  given  under  the  headings,  '•  The  Examination  and  Analysis 
of  Barley"  (p.  270),  and  "The  Examination  and  Analysis  of  Malt"  (p.  273). 

2  See  also  the  Saccharification  Test  (p.  274). 


250      Industrial  Organic  Analysis 

amount  of  maltose  formed  does  not  exceed  45  per  cent, 
of  the  starch  originally  present,  the  amount  of  maltose 
produced  in  a  given  time  may  be  taken  as  a  measure  of 
the  activity  of  the  diastase  solution  used.  The  actual 
method  described  is  due  to  Lintner.  In  the  first  place, 
the  preparation  of  soluble  starch  and  the  malt  extract 
will  be  described. 

Preparation  of  Soluble  Starch. — This  may  either  be 
obtained  ready-made  or  prepared  as  follows  :  100  grams 
of  purified  potato  starch  are  digested  with  500  c.c.  of 
dilute  hydrochloric  acid  of  specific  gravity  1-037,  a*  the 
ordinary  temperature,  for  7  days,  the  mixture  being 
stirred  daily.  The  mass  is  then  washed  repeatedly  by 
decantation,  first  with  tap  water  and  then  with  distilled 
water,  till  all  the  acid  has  been  removed.  When  the 
moist  starch  gives  no  acid  reaction  when  spread  on  blue 
litmus  paper,  a  few  drops  of  dilute  ammonia  are  added, 
and  the  starch  again  washed  by  decantation  until  all 
the  ammonia  has  been  removed.  It  is  then  collected  on 
a  Buchner  funnel,  sucked  as  dry  as  possible,  spread  on 
a  porous  plate  and  dried  at  a  gentle  heat  (about  30°). 
When  required  for  use,  the  pulverised  starch  is  dissolved 
in  boiling  water  to  make  a  2  per  cent,  solution  which,  on 
cooling,  should  be  perfectly  mobile,  and  not  gelatinous, 
as  would  be  the  case  with  a  solution  of  untreated  starch. 
Its  action  on  Fehling's  solution,  on  boiling,  should  be 
negligible. 

Preparation  of  Malt  Extract. — 25  grams  of  the  finely 
ground  malt  are  extracted  with  500  c.c.  of  distilled  water 
for  3  hours  at  21°,  agitating  the  mixture  from  time  to 
time.  The  whole  is  then  filtered,  the  first  100  c.c.  of  the 
filtrate  being  rejected.  The  extract  obtained  should  be 
perfectly  clear  and  bright. 

The  actual  determination  of  diastatic  activity  is 
carried  out  as  follows  :  Portions  of  10  c.c.  of  a  2  per  cent. 


Starch  and  'its  Products         251 

solution  of  soluble  starch  are  measured  out  in  eight 
carefully  cleaned  test  tubes,  which  are  placed  in  a  suitable 
stand  and  immersed  in  a  water  bath  kept  at  21°.  When 
the  starch  solution  has  reached  the  temperature  of  the 
bath,  o-i  c.c.  of  the  malt  extract  to  be  tested  is  measured 
into  the  first  tube,  0-2  c.c.  into  the  second,  0-3  into  the 
third,  and  so  on,  to  the  eighth  tube,  into  which  0-8  c.c. 
will  be  introduced.  In  examining  pale  malts,  i.e.,  malts 
which  usually  have  a  relatively  high  diastatic  power 
owing  to  the  lightness  of  the  curing,  it  is  recommended 
to  dilute  the  extract  obtained  as  directed  above  with 
an  equal  volume  of  water,  and  to  add  it  to  the  starch 
solution  in  the  tubes,  in  amounts  of  0-2  c.c.,  0-3  c.c., 
0-4  c.c.,  etc.  ;  in  this  way,  greater  accuracy  will  be 
attained.  The  tubes  are  allowed  to  remain  in  the  water 
bath  at  21°,  for  exactly  I  hour  from  the  time  the  extract 
has  been  added.  5  c.c.  of  mixed  Fehling's  solution  (see 
p.  139)  are  then  added  to  each  tube,  and  after  shaking 
the  tubes  are  heated  in  boiling  water  for  10  minutes  and 
allowed  to  stand  until  the  cuprous  oxide  has  settled. 
Usually  the  liquid  in  one  of  the  tubes  is  faintly  blue, 
showing  that  the  maltose  present  was  insufficient  in 
amount  to  effect  the  complete  reduction  of  the  cupric 
salt  present  in  the  5  c.c.  of  Fehling's  solution,  while  the 
liquid  in  the  next  tube  of  the  series  is  yellow,  owing  to 
over  reduction.  In  this  way  the;  amount  of  malt  extract 
which  is  just  sufficient  to  produce,  in  the  given  time,  the 
quantity  of  maltose  to  reduce  5  c.c.  of  Fehling's  solution 
may  be  estimated.  Thus,  if  tube  No.  2  is  as  much 
under  reduced  as  tube  No.  3  is  over  reduced,  then  the 
amount  of  malt  extract  sufficient  to  cause  complete 
reduction  may  be  taken  at  0-25  c.c.  (i.e.,  if  tube  No.  2 
contained  0-2  c.c.,  and  tube  No.  3  contained  0-3  c.c.  of 
the  extract).  If  the  liquid  in  one  of  the  tubes  is  neither 


252      Industrial  Organic  Analysis 

blue  nor  yellow,  then  the  maltose  formed  therein  will 
have  been  just  sufficient  to  cause  complete  reduction. 

Complete  reduction  by  o-i  c.c.  of  malt  extract  corre- 
sponds to  a  Lintner  value  of  100.  If  X  is  the  quantity 
of  malt  extract  required  for  complete  reduction,  in 
cubic  centimetres,  then  the  diastatic  activity  of  the  malt, 

A  _  o-i  x  100 

~X~ 

From  the  figure  thus  found,  it  is  usual  to  deduct  1-5, 
to  allow  for  the  reducing  sugars  present  in  the  malt 
extract.  If  the  extract  has  been  diluted  with  an  equal 
quantity  of  water,  then  half  the  volume  actually  added 
must  be  used  for  the  purposes  of  the  calculation. 

Diastatic  activities  of  over  80  are  considered  high, 
under  50  low. 

The  examination  of  malt  is  further  described  below. 
(See  p.  273.) 

FLOUR. 
INTRODUCTORY. 

The  most  important  of  the  starch-containing  food 
materials  are  derived  from  the  grain  of  the  cereals,  such 
as  wheat,  rye,  maize,  barley,  etc.,  after  removal  of  the 
chaff  (paleae  and  glumes)  by  threshing.  The  coarsely 
ground  grain  is  known  as  meal,  the  finely  ground  grain 
as  flour.  Special  attention  is  paid  to  the  examination 
of  wheaten  flour,  as  this  is  the  most  important  of  the 
flours  consumed  in  this  country,  as  well  as  being  the 
dearest,  and  therefore  the  most  likely  to  be  adulterated 
with  other  flours. 

The  following  table  contains  results  published  by  the 
United  States  Department  of  Agriculture,  showing  the 
average  composition  (percentage)  of  the  more  important 
cereals  ;  a  typical  analysis  of  potato  is  appended. 

The  products  obtained  from  the  grain  of  wheat  on 
milling  may  be  divided  into  the  flour  proper,  by  which 
is  understood  the  contents  of  the  parenchymatous  cells 


Starch  and  its  Products 


253 


Description  of  Grain. 

Mois- 
ture. 

Nitro- 
gen 
x  6-25. 

Crude 
Fibre. 

Ash. 

Carbo- 
hydrates 
other 
than 
Crude 

Js 

ti3   t*^ 

si 

Weight 
of  100 
Kernels 
in 

^"o 

Grams. 

Fibre. 

»- 

Typical     unhulled 

barley 

10  85 

II-O 

3-85 

2-5 

69-55 

2.25 

— 

Typical   American 

^  maize 

10-75 

10-0 

J'75 

1-5 

71-75 

4-25 

38-0 

Typical  rye  . 

10-5 

12-25 

2-1 

1-9 

7^75 

2'5 

Typical     unhulled 

oats 

IO-0 

12-0 

12-0 

3'4 

58-0 

4'5 

3-° 

Typical      rice, 

unhulled  . 

10-5 

7'5 

9-0 

4-0 

67-4 

1-6 

3-0 

Typical      rice, 

polished   . 

12-4 

7-5 

0-4 

0-5 

78-8 

0-4 

2-2 

Typical  wheat 

3-85 

12-25 

2-4 

1-75 

71-25 

i-75 

3-85 

Potato 

74-7 

2-O 

1-4 

i-o 

20-7 

0-2 

of  the  endosperm,  and  the  offal,  which  includes  the  bran, 
or  husk,  the  germ,  i.e.,  the  embryo  plant,  and  fluffy  or 
fibrous  matter  derived  from  the  cell  walls  of  the  parenchy- 
matous  endosperm  ;  "  sharps,"  or  "  middlings,"  consist 
of  finely  divided  bran  with  particles  of  germ.  The  com- 
mercial valuation  of  flours  varies  according  to  the  amount 
of  offal  which  they  contain,  the  highest  grade  being  the 
so-called  "  patent  flour  "  which,  owing  to  its  freedom 
from  offal,  is  the  whitest  in  colour  and  most  homogeneous 
product ;  it  is  appreciated  by  bakers  on  account  of  the 
amount  of  water  which  it  will  absorb  and  the  appearance 
of  the  loaf  which  it  produces,  though  as  regards  nutritive 
value  it  is  not  necessarily  superior  to  the  lower  grades. 
"  Wholemeal,"  or  "  Graham,"  flour  is  the  product  of  grind- 
ing of  the  entire  wheat  grain,  including  the  husk  and  germ ; 
it  should  have  the  same  composition  as  the  wheat  grain 
itself  ;  containing  the  whole  of  the  offal,  it  is  considerably 
darker  in  colour  than  patent  flour,  and  not  being  sub- 
jected to  any  sifting  process,  it  contains  branny  particles 
which  are  distinctly  visible.  The  so-called  "  entire 
wheat  flour,"  or  fine  meal,  is  the  product  obtained  by 
removing  a  portion  of  the  bran  and  finely  grinding  the 
rest  of  the  grain  ;  the  texture  is  somewhat  coarser  than 


254      Industrial  Organic  Analysis 

that  of  ordinary  or  patent  flour,  the  colour  and  general 
appearance  varying  somewhat,  according  to  the  amount 
of  offal  removed  and  the  nature  of  the  wheat  from  which 
it  is  milled.  "  Entire  "  flour  usually  contains  a  portion  of 
the  germ.  "  Standard,"  or  80  per  cent,  flour,  may  be 
classed  as  an  "entire"  flour,  representing  80  per  cent,  of 
the  wheat  grain,  and  containing  the  whole  of  the  germ. 

"  Households  grade  "  is  the  commercially  lower  grade 
of  flour  which  is  produced  besides  the  patent  grade  in 
the  modern  process  of  roller  milling  ;  it  is  darkish  in 
colour  and  contains  a  small  amount  of  fine  branny 
particles  ;  from  the  baker's  point  of  view,  it  is  inferior 
to  patent  flour  in  bread  making.  "  Baker's  grade  "  and 
"  clear  grade  "  are  terms  applied  to  similar  grades  in 
America.  "  Straight  run/'  or  "  straight  grade,"  may  be 
regarded  as  a  mixture  of  the  patent  and  household  grades. 

Special  flours  may  be  prepared  from  any  of  the  above 
grades,  usually  with  a  view  to  improving  the  nutritive 
value.  They  may  contain  finely  powdered  bran,  lentil 
flour  or  banana  meal.  Germ  flours  contain  added  germ 
(usually  cooked). 

The  above  is  an  abridged  version  of  Dr.  Hamill's 
classification  of  the  various  flours  on  the  market,  included 
in  his  report  to  the  Local  Government  Board  on  the 
nutritive  value  of  bread  made  from  different  varieties  of 
wheat  flour.  As  is  pointed  out  by  Dr.  Hamill,  it  is  very 
difficult  to  set  up  analytical  standards  for  denning  any 
particular  grade  of  flour,  owing  to  the  difference  in  com- 
position of  wheats  from  different  sources.  Comparisons 
of  the  composition  of  different  grades  of  flour  can  only 
be  of  value  when  these  are  milled  from  the  same  wheat. 

The  following  table,  compiled  by  the  United  States 
Department  of  Agriculture,  shows  the  variations  in  the 
composition  of  wheats  from  different  sources  :— 

Per  cent. 

Water         .         .         .         .         .         7  to  14 
Proteins      .          .          .          .          .          8      17 


Dry  gluten           ....  2 

Carbohydrates     ....  65 

Ether  extract  (chiefly  fat)     .          .  0-28 

Ash   .         .         .         .         .         •  i'4 


14 
76 

2'5 
2-3 


Starch  and  its  Products         255 

A  number  of  analyses  of  various  flours  and  other 
milling  products  obtained  from  different  wheat  mixtures 
are  given  in  Dr.  HarrmTs  report.  From  these  and  from 
the  figures  given  in  the  next  table,  it  will  be  seen  that 
the  presence  of  bran  in  the  flour  tends  to  raise  the  con- 
tent of  mineral  matter  and  crude  fibre.  The  presence 
of  germ  tends  to  raise  the  content  of  fat  and  protein,  but 
usually  these  increases  are  comparatively  insignificant. 
"  Germ  flours  "  may,  however,  contain  as  much  as  4  to  5 
per  cent,  of  additional  protein  due  to  added  germ. 

The  following  table  from  Wynter  Bryth's  "  Foods,  their 
Composition  and  Analysis,"  shows  the  compositions  of 
various  flours  of  different  grades,  the  original  wheat 
from  which  these  are  milled,  and  of  the  bran.  The 
superiority  of  grade  and  freedom  from  bran  decrease 
from  Nos.  o  to  9  :— 


Description    of    Flour 
of  Varying  Grades 
as  obtained  from 
Steel  Roller  Mills  and 
Bran. 

Per  Cent. 
Yield  on 
Original 
Wheat. 

Water. 

Insol- 
uble 

Nitro- 
genous 
Matter. 

Soluble 
Nitro- 
genous 
Matter.l 

Fat. 

Carbo- 
hydrates 
other 
than 
Crude 
Fibre. 

Crude 
Fibre. 

Ash. 

Original  wheat 

I3-37 

10*69 

2-93 

1-98 

80-41 

1-90 

2-09 

Flour  No.  o 

6-0 

12*56 

8-38 

3-o6 

0-83 

87-26 

trace. 

0-47 

„         „     2 

6-0 

12-48 

8-87 

2'95 

0-97 

86-69 

M 

0-52 

„    6 

4-0 

12-39 

9-38 

3'oo 

i'i7 

85-87 

0'02 

0*56 

It              M       8 

5  to  6 

11-72 

10-06 

2-72 

T'30 

75-90 

0*06 

o'8i 

to 

to 

to 

to 

to 

to 

to 

12-41 

I4'34 

3-22 

3-5I 

84-55 

1-03 

2-23 

».        „    9 

yo 

10-64 

15-02 

2-55 

4-02 

74-20 

1*53 

2-66 

Fine  bran 

i6'o 

«'35 

13-50 

3-o6 

4'54 

63-64 

8-71 

6-55 

Coarse  bran 

2-0 

12-37 

13*44 

3-17 

3-46 

62-13 

9'79 

8-01 

Soluble  nitrogen,  estimated  by  Weinwurm,  according  to  the  method  described  below. 


The  crude  fibre  is  the  cellulose  vegetable  tissue  which 
is  insoluble  in  dilute  boiling  acid  and  alkali  ;  it  cannot 
be  looked  on  as  a  definite  constituent,  the  amount  found 
varying  somewhat  according  to  the  method  of  estimation. 
On  examination  of  the  ash,  or  mineral  matter,  the 
principal  bases  which  will  be  found  are  potash,  lime  and 
magnesia,  and  the  principal  acid,  phosphoric  acid.  The 
nitrogenous  constituents  of  wheat  flour  may  be  divided 


256      Industrial  Organic  Analysis 

into  the  gluten  and  the  soluble  nitrogen  compounds  ; 
they  are  of  importance,  not  only  on  account  of  their 
nutritive  value,  but  also  on  account  of  the  texture  which 
they  impart  to  the  bread.  Gluten  consists  mainly  of 
two  proteins,  gliadin  and  glutenin.  The  former  is  a  soft, 
sticky  substance  which  can  be  pulled  out  into  threads  ; 
it  may  be  separated  from  the  other  constituents  of  flour 
by  extraction  with  70  per  cent,  alcohol,  in  which  it  is 
soluble,  and  by  precipitation  from  this  solution  by  the 
addition  of  an  aqueous  solution  of  sodium  chloride. 
Though  soluble  in  pure  water,  gliadin  is  not  dissolved 
when  flour  is  treated  with  water,  owing  to  the  presence 
of  mineral  salts.  Glutenin  is  distinguished  by  its 
solubility  in  very  dilute  alkali  solution  ;  it  is  of  a  firmer 
consistency  than  gliadin  ;  the  latter  imparts  tenacity  to 
the  gluten,  and  on  this  account,  bakers  usually  prefer 
a  flour  containing  a  high  percentage  of  gliadin.  The 
proteins  other  than  gluten  include  a  globulin,  an  albumin 
and  a  proteose  ;  these  are  probably  of  greater  nutritive 
value  than  the  gluten. 

THE  EXAMINATION  AND  ANALYSIS  OF  FLOUR. 

As  has  been  previously  pointed  out,  the  commercial 
value  of  wheat  flour  depends  on  its  colour,  texture,  and 
the  amount  and  nature  of  the  gluten  present.  The 
highest  grade  flours  should  be  practically  white,  showing 
only  the  faintest  tinge  of  yellow  ;  they  should  possess 
a  sweet  smell  and  be  free  from  branny  particles  and 
acidity.  By  the  tests  and  analytical  methods  now  to  be 
described,  the  quality  of  a  flour  may  be  gauged  to  a 
certain  extent ;  a  great  deal,  however,  depends  on  the 
appearance  of  the  flour  and  the  loaf  which  it  will  make. 
The  detection  of  added  foreign  matter  and  foreign  flours 
will  also  be  dealt  with. 

Gluten  Test. — In  this  test,  the  gluten  is  separated  from 
the  other  constituents  of  the  flour  by  mechanical  means, 
examined  and  estimated.  About  30  grams  of  flour  are 
made  into  a  stiff  dough  with  about  12  to  15  c.c.  of  water, 
and  allowed  to  stand  for  an  hour.  The  mass  is  then 


Starch  and  its  Products         257 

carefully  kneaded  in  a  stream  of  running  water,  over 
muslin  until  all  the  starch  has  been  removed.  The 
fresh  gluten  thus  obtained  should  only  have  a  faint  yellow 
tinge  and  should  be  of  such  a  consistency  that  it  can  be 
pulled  out  into  threads  ;  the  gluten  from  English  wheat 
is  usually  very  soft  and  sticky,  having  less  elasticity  than 
gluten  from  other  wheats.  A  dark  and  viscous  gluten 
indicates  the  presence  of  rye  flour  ;  most  of  the  other 
common  flours  also  give  more  or  less  coloured  glutens, 
thus,  that  from  barley  is  non- viscous  and  dirty  reddish 
brown,  from  oats,  dark  yellow,  from  maize  yellowish  and 
non-elastic,  from  leguminous  flours  such  as  those  of  the 
bean  or  pea,  greyish  red  to  green.  A  grey  or  reddish 
gluten  may  also  be  obtained  from  inferior  wheat  flour. 
After  washing,  the  gluten  is  left  under  water  for  an  hour, 
after  which  the  excess  of  water  is  removed  as  completely 
as  possible  by  pressing  with  the  hands,  and  the  moist 
gluten  weighed.  It  is  then  dried  at  100°  till  constant  in 
weight,  which  may  take  24  hours  or  more.  A  good 
wheat  flour  will  contain  from  20  to  40  per  cent,  of  moist 
gluten  and  about  10  to  18  per  cent,  of  gluten  dried  at 
100°.  When  heated  to  150°,  good  gluten  swells  and 
assumes  the  appearance  of  bread. 

The  above  test,  although  useful  for  valuating  flour, 
cannot,  of  course,  be  looked  on  as  an  accurate  analytical 
process  ;  the  gluten  obtained  will  contain  small  per- 
centages of  non-gluten  proteins,  mineral  matter,  fat, 
starch,  fibre,  etc. 

Total  Proteins. — The  total  nitrogen  is  determined  in 
about  2  to  3  grams  of  the  original  flour  by  the  Kjeldahl 
method,  as  described  in  Chapter  I.,  p.  22.  The  nitrogen 
present  as  nitrates,  which  exist  in  small  quantity  in 
wheat,  will,  of  course,  not  be  converted  into  ammonia 
and  estimated  by  this  method  (see  p.  26).  The  average 

I.O.A.  17 


258      Industrial  Organic  Analysis 

nitrogen  content  of  wheat  proteins  being  17-6  per  cent., 
the  nitrogen  found  should  be  multiplied  by  5-68,  in  order 
to  give  the  total  proteins.  For  many  other  cereal  pro- 
teins, the  factor  6-25  gives  a  more  accurate  result. 
Reference  to  the  last  table  will  show  that  there  is  a 
tendency  for  the  protein  content  to  increase  in  passing 
from  the  high-grade  flours,  through  the  lower-grade 
flours,  to  bran.  In  this  connection,  however,  it  must  be 
remembered  that  the  protein  content  of  wheat  from 
different  sources  varies  within  fairly  wide  limits  (see  the 
previous  tables).  As  was  pointed  out  above,  the  inclu- 
sion of  the  germ  in  the  flour  will  not  materially  raise  the 
protein  content,  though  if  much  extra  germ  be  added,  as 
in  the  case  of  the  so-called  germ  flours,  the  protein 
content  may  be  raised  by  as  much  as  4  to  5  per  cent, 
above  the  normal. 

According  to  the  United  States  standard  of  purity, 
wheat  flour  must  contain  not  less  than  1-25  per  cent,  of 
nitrogen. 

Gliadin. — This  may  be  estimated  by  extracting  the 
flour  for  2  hours  with  70  per  cent,  alcohol,  filtering,  and 
determining  the  nitrogen  in  an  aliquot  portion  of  the 
filtrate.  For  conversion  of  the  nitrogen  found  into 
gliadin  the  factor  5-68  should  be  used. 

In  a  good  flour  the  gliadin  should  constitute  about 
60  per  cent,  or  more  of  the  total  gluten. 

Soluble  Nitrogenous  Matter. — This  is  estimated  by 
Weinwurm  as  follows  :  10  grams  of  the  material  are 
treated  with  200  c.c.  of  hot  water  and  0-5  c.c.  of  acetic 
acid,  and  the  whole  is  warmed  on  a  water  bath  for 
20  minutes.  The  solution  is  cooled,  made  up  to  500  c.c., 
and  filtered  ;  the  soluble  nitrogen  is  then  estimated  in 
50  c.c.  of  the  filtrate,  which  should  be  e.vaporated  nearly 


Starch  and  its  Products         259 

to  dryness  before  adding  the  sulphuric  acid  for  the 
Kjeldahl  estimation.  Our  knowledge  regarding  the 
relative  digestibility  of  the  various  proteins  of  the  cereals 
is  somewhat  limited,  though  it  may  be  presumed  that 
the  soluble  nitrogenous  matter  as  estimated  by  the  above 
method  has  greater  nutritive  value  than  the  proteins 
of  the  gluten.  Several  other  methods  have  been  devised 
for  estimating  the  "  digestible  nitrogenous  matter." 

Water. — i  to  3  grams  of  the  flour  are  weighed  out 
between  watch  glasses,  and  dried  in  a  steam  oven 
till  no  further  loss  in  weight  takes  place. 

In  the  United  States  and  Canada,  the  maximum  limit 
for  water  in  flour  is  fixed  at  13  per  cent.  An  unduly 
large  proportion  of  water  impairs  the  keeping  qualities 
of  the  flour,  besides,  of  course,  lessening  its  nutritive 
value. 

Ash. — About  5  grams  of  the  flour  are  burnt  in  a 
crucible,  in  a  muffle  furnace,  and  the  residual  ash  weighed. 
An  alternative  method  which  is  sometimes  used  is  to 
mix  the  flour  with  powdered  ammonium  nitrate,  heating 
the  mixture  carefully  and  withdrawing  the  flame 
directly  fusion  commences.  In  this  way,  the  flour  may 
be. burnt  up  quickly,  without  the  use  of  the  muffle  furnace. 
A  corresponding  quantity  of  ammonium  nitrate  should 
be  heated  separately,  and  the  residue  which  it  leaves,  if 
any,  determined  and  deducted  from  the  ash  found  in  the 
actual  estimation. 

The  amount  of  ash  or  mineral  matter  present  in  the 
patent  and  other  higher  grades  of  flour  is  usually  under 
0-5  per  cent.  ;  households  and  "  standard,"  or  80  per 
cent,  flour,  generally  contain  between  0-5  and  0-9  percent, 
of  mineral  matter,  while  wholemeal  may  contain  between 
1-5  and  2  per  cent.,  or  even  more.  From  a  study  of  the 
table  on  p.  255,  it  will  be  obvious  that  the  greater  the 
proportfon  of  bran  or  offal  left  in  the  flour,  the  more 

17—2 


260      Industrial  Organic  Analysis 

mineral  matter  will  the  latter  contain.  According  to  the 
United  States  standard  of  purity,  wheat  flour  should  not 
yield  more  than  i-o  per  cent,  of  ash.  If  the  amount  of 
ash  found  seems  abnormally  high  in  relation  to  the  quality 
of  the  flour,  it  is  possible  that  mineral  matter  has  been 
added  with  a  view  to  improving  its  appearance.  Acid 
potassium,  magnesium  or  calcium  phosphates  may  some- 
times be  added  as  "  improvers  "  in  the  proportion  of 
about  0-5  per  cent,  of  the  flour.  If  such  additions  are  sus- 
pected, the  ash  should  be  preserved  for  further  examina- 
tion. Alum  and  copper  sulphate  may  also  be  used  as 
improvers,  but  the  quantities  added  will  be  too  small  to 
be  detected  by  any  increase  in  the  proportion  of  the  ash. 
Self-raising  flours  usually  contain  added  sodium  bicar- 
bonate and  acid  calcium  phosphate,  these  constituents 
sometimes  being  added  in  the  proportion  of  about  3  per 
cent,  of  the  flour.  The  detection  of  added  mineral 
matter  will  be  dealt  with  later. 

Starch. — This  is  best  estimated  by  O'Sullivan's  method, 
described  on  pp.  236  et  seq. 

Acidity. — 20  grams  of  the  flour  are  shaken  with  200  c.c. 
of  water  at  intervals,  for  2  hours,  the  mixture  is  filtered, 
and  50  c.c.  of  the  filtrate  titrated  with  decinormal 
sodium  hydroxide  solution,  using  phenol  phthalein  as 
indicator. 

According  to  Wynter  Blyth,  normal  wheat  should 
show  an  acidity  corresponding  to  not  more  than  0-16  to 
0-25  per  cent,  of  lactic  acid.  The  test  is  useful  for 
detecting  the  presence  of  unsound  wheat  in  flour. 

Fat  (Ether  Extract). — This  is  determined  by  extracting 
about  3  grams  of  well-dried  flour  with  ether  in  a  Soxhlet 
apparatus.  (See  p.  66.)  The  extracted  material 
should  be  preserved  for  the  estimation  of  crude  fibre. 
When  the  extraction  is  complete,  the  ethereal  solution 
is  evaporated  in  a  tared  flask,  the  residue  dried  at  105°, 
and  weighed. 


Starch  and  its   Products         261 

Although,  generally  speaking,  it  may  be  said  that  the 
greater  the  proportion  of  bran  in  the  flour  the  higher  the 
fat  percentage  (see  the  table  on  p.  255),  the  latter  can 
hardly  be  taken  as  an  index  of  the  quality  of  the  flour, 
chiefly  owing  to  the  variations  in  the  fat  content  of 
flours  from  different  sources. 

Crude  Fibre. — The  following  method  is  recommended 
by  the  American  Association  of  Official  Agricultural 
Chemists  for  the  estimation  of  crude  fibre  in  grain, 
flour,  and  bye-product  cattle  foods,  such  as  oil  cake,  etc.  : 
2  grams  of  the  material  are  extracted  with  ether,  or  the 
extracted  material  from  the  fat  determination  may  be 
used.  The  fat  free  material  is  placed  in  a  500  c.c.  flask, 
and  200  c.c.  of  boiling  1-25  per  cent,  sulphuric  acid  are 
added  ;  the  flask  is  connected  with  a  reflux  condenser  by 
means  of  a  rubber  stopper ;  boiling  is  commenced  at 
once  and  continued  for  30  minutes.  A  current  of  air 
passing  through  the  liquid  may  serve  to  reduce  frothing. 
The  mixture  is  filtered,  and  the  residue  washed  with 
boiling  water  until  the  washings  are  no  longer  acid  ;  it 
is  then  rinsed  back  into  the  same  flask  with  200  c.c.  of  a 
boiling  1-25  per  cent,  solution  of  sodium  hydroxide,  free 
or  nearly  so,  from  carbonate  ;  boiling  is  commenced  at 
once,  and  continued  for  30  minutes,  in  the  same  manner  as 
directed  above  for  the  treatment  with  acid.  The  insoluble 
residue  is  filtered  off  on  a  Gooch  crucible  and  washed 
with  boiling  water  until  the  washings  are  neutral,  dried 
at  110°  and  weighed  ;  it  is  then  incinerated  completely  ; 
the  loss  in  weight  thus  occasioned  represents  crude  fibre. 

The  filter  used  for  the  first  filtration  may  be  of  linen, 
glass  wool,  asbestos  or  any  convenient  form  giving  clear 
and  reasonably  rapid  filtration.  The  strength  of  the 
acid  and  alkali  solutions  employed  should  be  verified  by 
titration,  and  not  merely  by  the  specific  gravity. 

According  to  the  United  States  standard  of  purity, 


262      Industrial  Organic  Analysis 

wheat  flour  should  not  contain  more  than  0-5  per  cent, 
of  crude  fibre.  As  will  be  seen  from  the  table  on  p.  255, 
the  crude  fibre  in  the  highest  grades  of  wheat  flour  only 
amounts  to  traces,  and  increases  with  the  proportion  of 
bran  present.  Provided  that  a  perfectly  uniform  method 
of  estimation  is  adopted,  the  percentage  of  crude  fibre  is, 
perhaps,  the  most  reliable  index  of  the  amount  of  branny 
matter  in  the  flour. 


THE  DETECTION  AND  ESTIMATION  OF  FOREIGN  MATTER 
AND  ADULTERANTS  IN  WHEAT  FLOUR. 

Under  this  heading  the  detection  of  added  mineral 
matter,  foreign  flours  or  starches,  and  flour  damaged  by 
fungi,  such  as  ergot,  are  considered.  Added  mineral 
matter  cannot,  as  a  rule,  be  detected  with  any  certainty 
by  an  increase  in  the  ash  content  of  the  flour,  owing  to 
the  variations  in  ash  content  of  flours  of  differing  fineness 
and .  origin  ;  for  this  purpose,  the  special  methods 
described  below  must  be  adopted.  Addition  of  foreign 
flours,  etc.,  may  sometimes  be  detected  by  the  use  of  the 
microscope,  though  for  this  task  considerable  experience 
is  often  necessary.  In  some  cases,  however,  chemical 
methods  are  available. 

Alum. — Alum  is  sometimes  added  to  bad  or  slightly 
damaged  flour  by  the  miller  or  the  baker  in  order  to 
improve  its  appearance  ;  although  it  is  only  added  in 
very  small  quantities,  which  can  hardly  be  injurious  to 
health,  its  use  is  prohibited  by  law  in  England.  Two 
methods  are  available  for  the  detection  of  alum  in 
flour,  i.e.,  the  logwood  method  and  the  chloroform 
method  ;  the  latter  method  also  allows  of  the  detection 
of  mineral  additions  other  than  alum. 

Logwood  Method. — Tincture  of  logwood,  which  should 
always  be  freshly  made,  is  prepared  as  follows  :  J  gram 
of  fine  logwood  chips,  preferably  cut  fresh  from  the  log, 
is  macerated  for  10  hours  in  15  c.c.  of  alcohol ;  10  c.c.  of 
the  solution  are  poured  off  and  mixed  with  150  c.c.  of 
water  and  10  c.c.  of  a  saturated  solution  of  ammonium 


Starch  and  its  Products         263 

carbonate.  The  test  should  be  carried  out  immediately 
after  the  addition  of  the  ammonium  carbonate  ;  50  grams 
of  the  flour  are  made  into  a  thin  paste  with  water,  a 
few  drops  of  the  logwood  solution  are  added,  and  the 
mixture  allowed  to  stand  for  several  hours.  If  alum  is 
present  a  lavender  blue  lake  will  be  produced.  The 
colour  should  persist  when  the  sample  is  placed  in  an 
oven  at  100°  for  2  hours.  If  only  i  part  of  alum  in  1,000 
be  present,  the  flour  becomes  pink  instead  of  lavender. 
In  the  case  of  a  negative  result  being  obtained  by  this 
test,  the  absence  of  alum  may  be  inferred.  If  a  positive 
result  is  obtained,  it  is  best  to  confirm  it  by  the  chloro- 
form method  in  order  to  make  sure  that  the  coloration  is 
not  due  to  adventitious  clayey  matter  from  the  mill- 
stones ;  in  the  modern  process  of  milling,  however,  in 
which  steel  rollers  are  used,  the  latter  source  of  contami- 
nation is  avoided.  Moreover,  the  logwood  test  is  only 
obtained  in  the  presence  of  aluminium  compounds 
soluble  in  water. 

Chloroform  Method. — This  method  also  allows  of  the 
detection  of  mineral  matter  other  than  alum  ;  as 
mentioned  above,  alum  and  copper  sulphate  are  only 
added  in  very  small  quantities,  their  function  being  to 
increase  the  whiteness  of  the  flour  ;  their  use  is  forbidden 
by  law,  and  it  is  but  rarely  that  they  will  be  found  in 
commercial  samples  of  flour.  Substances  such  as  acid 
potassium,  magnesium  or  calcium  phosphates,  notably 
the  latter,  are  used  in  larger  quantities,  their  function 
likewise  being  to  make  lower  grades  of  flour  appear  equal 
to  the  highest  grades  ;  they  are  said  to  "  make  the  flour 
bake  whiter,  while  the  bread  is  improved  in  boldness, 
texture,  crust,  etc."  Dr.  Hamill,  in  his  report  to  the 
Local  Government  Board  on  the  bleaching  of  flour 
and  the  addition  of  so-called  "  improvers "  to  flour, 
states  that  he  has  seen  acid  calcium  phosphate  added  to, 
and  intimately  mixed  with,  flour  in  the  proportion  of 
about  0-45  per  cent,  of  the  finished  flour.  Curtel 


264      Industrial  Organic  Analysis 

(Annales  des  Falsifications,  1910,  p.  302)  describes  a 
preparation  sold  under  the  name  "  Blanc  Flour/'  con- 
sisting essentially  of  acid  calcium  phosphate,  which  the 
makers  recommend  should  be  added  in  the  proportion 
of  i  per  cent,  of  the  flour.  Considerable  quantities  of 
calcium  sulphate  may  be  introduced  if  inferior  grades  of 
calcium  phosphate  are  used,  especially  in  the  case  of 
self-raising  flours,  which  are  now  usually  prepared  by 
adding  calcium  acid  phosphate  and  sodium  bicarbonate 
to  the  flour.  As  mentioned  above,  the  amount  of  mineral 
matter  added  in  such  cases  is  usually  large  enough  to  be 
detected  by  an  examination  of  the  ash  of  the  flour.  In 
a  report  to  the  Local  Government  Board  it  is  recom- 
mended that  bakers  and  millers  should  insist  on  a  calcium 
phosphate  containing  under  10  per  cent,  of  calcium 
sulphate.  If  a  conspicuous  excess  of  the  latter  be  found, 
prosecution  under  the  Sale  of  Foods  and  Drugs  Act  of 
1875  might  be  considered. 

The  detection  of  added  mineral  matter  by  the  chloro- 
form method  may  be  carried  out  as  follows  :  200  grams 
of  flour  are  shaken  in  a  separating  funnel  with  sufficient 
chloroform  to  give  a  perfectly  liquid  mixture,  and  allowed 
to  stand  overnight.  The  chloroform  is  of  sufficient 
density  to  float  the  organic  constituents  of  the  flour, 
while  the  mineral  matter  will  sink  to  the  bottom,  without 
being  dissolved.  The  solid  matter  which  has  collected 
at  the  bottom  is  carefully  removed  through  the  stopcock, 
shaken  a  second  time  with  a  little  more  chloroform,  and 
when  it  has  subsided  again,  transferred  to  a  watch  glass 
so  that  the  chloroform  may  evaporate.  The  residue  is 
treated  with  a  small  quantity  of  water,  the  solution  is 
separated  from  the  insoluble  portion  and  allowed  to 
evaporate,  when  the  alum  will  separate>  if  present,  in 
the  form  of  octahedral  crystals,  which  may  be  identified 
by  examination  under  a  low-power  microscope.  The 
crystals  may  be  dissolved  in  water,  and  identified  as  alum 
by  the  usual  tests.  The  residue  insoluble  in  water  may 


Starch  and  its  Products         265 

be  examined  under  the  microscope,  and  subjected  to  a 
qualitative  analysis  for  metals  and  acid  radicles  ;  small 
amounts  of  copper  salts,  either  in  the  soluble  or  insoluble 
portion,  may  be  detected  by  their  delicate  reaction  with 
potassium  ferrocyanide. 

For  the  detection  of  calcium  phosphate  in  flour,  Curtel 
(loc.  tit.)  recommends  the  following  method,  which,  in 
principle,  is  practically  identical  with  the  foregoing  : 
5  grams  of  flour  are  shaken  with  40  to  50  c.c.  of  carbon 
tetrachloride,  the  mixture  is  centrifuged,  and  the  flour 
and  the  carbon  tetrachloride  are  separated  from  the 
sediment ;  the  latter  is  dissolved  in  a  little  nitric  acid, 
and  tested  with  a  small  quantity  of  ammonium  molyb- 
date  solution  ;  if  the  flour  contained  added  calcium 
phosphate,  a  copious  precipitate  will  be  obtained  ;  the 
small  amount  of  sediment  obtained  from  unadulterated 
flours,  representing  natural  impurities,  will  produce  no 
precipitate.  The  use  of  the  centrifuge,  both  in  this  and 
the  forgeoing  method,  greatly  facilitates  the  separation 
of  the  mineral  sediment  from  the  flour.  The  advantage 
of  the  two  methods  just  described  lies  in  the  fact  that 
practically  only  the  mechanically  admixed  matter  is 
separated  ;  moreover,  by  the  use  of  such  liquids  as 
chloroform  or  carbon  tetrachloride,  interaction  between 
the  added  mineral  matter  and  the  natural  constituents 
of  the  wheat  is  avoided,  such  as,  for  example,  would 
occur  between  alum  and  the  wheat  phosphates  in 
presence  of  water. 

Detection  of  Flour  Damaged  by  Moulds. — After  bad 
seasons,  flour  is  liable  to  contain  matter  which  has  been 
damaged  by  moulds,  especially  ergot.  For  the  detection 
of  starch  which  has  been  affected  by  moulds,  a  minute 
quantity  of  the  flour  may  be  mounted  in  a  drop  of 
glycerol,  and  examined  under  a  low-power  microscope  ; 


266      Industrial  Organic  Analysis 

a  drop  of  aniline  violet  solution  is  then  allowed  to  diffuse 
into  the  mounting  medium,  under  the  cover  glass  ; 
granules  which  have  been  damaged  by  mould  of  any  kind 
will  take  up  the  colour  intensely. 

Chemical  Method  for  Detecting  Ergot. — The  following 
method  is  given  by  Leffmann  and  Beam  :  10  grams  of 
flour  are  macerated  for  about  30  minutes  with  a  mixture 
of  20  c.c.  of  ether  and  10  drops  of  dilute  sulphuric  acid 
(i  to  5  parts  water)  ;  the  mixture  is  filtered  and  the 
residue  is  washed  with  ether  until  the  filtrate  measures 
15  c.c.  This  filtrate  is  shaken  up  with  5  drops  of  a 
saturated  solution  of  sodium  bicarbonate.  The  green 
chlorophyll  remains  in  the  ether  ;  the  bicarbonate  solu- 
tion will  remain  clear  if  the  flour  be  derived  from  sound 
grain,  but  becomes  deep  violet  if  ergot  be  present. 

The  sale  of  ergotised  flour  as  good  flour  would  con- 
stitute an  offence  under  the  Sale  of  Foods  and  Drugs  Act. 

The  Detection  of  Foreign  Flours  in  Wheat  Flour. — 
According  to  Wynter  Blyth  the  following  have  been 
fraudulently  added  to  flour  :  rye,  rice  and  barley  meals, 
flours  from  various  leguminosae,  such  as  the  bean  or  pea, 
linseed  meal,  buckwheat,  potato  and  some  other  starches. 
Such  additions  are  stated  to  be  rare  in  this  country,  the 
foreign  flours  most  likely  to  be  met  with  being  those  of 
rice  and  the  potato  ;  adulterations  of  this  nature  are, 
naturally,  most  frequently  to  be  met  with  in  practice  in 
times  of  scarcity  of  wheat. 

General  Tests. — Vogel  extracts  the  suspected  flour  with 
70  per  cent,  alcohol,  to  which  5  per  cent,  of  hydrochloric 
acid  has  been  added  ;  if  the  flour  is  derived  from  pure 
wheat  or  rye,  the  alcohol  remains  colourless,  but  takes 
up  a  yellow  colour  if  either  barley  or  oats  be  present, 
orange  yellow  in  presence  of  pea-flour,  and  purple  red  or 
blood  red  in  presence  of  mildewed  or  ergotised  wheat, 
respectively. 


Starch  and  its  Products         267 

The  conclusions  which  may  be  drawn  from  the 
appearance  of  the  gluten,  are  given  under  the  gluten  test. 
(See  p.  256.)  Special  methods  for  the  detection  of  foreign 
flours  will  now  be  described. 

Potato  Starch. — Donne's  test  for  the  detection  of 
potato  starch  in  wheat  flour  is  carried  out  as  follows  : 
the  flour  is  examined  in  a  thin  layer,  mounted  in  water, 
in  the  usual  way,  under  the  microscope  ;  then,  while  the 
starch  is  still  under  observation,  a  weak  solution  of 
potassium  hydroxide  is  allowed  to  diffuse  under  the  cover 
glass.  Under  the  influence  of  the  alkali,  potato  starch 
granules  will  begin  to  swell  till  they  reach  4  to  5  times 
their  original  volume,  while  wheat  starch  is  scarcely 
affected.  In  order  to  render  this  test  sharper,  it  may  be 
combined  with  Lecanu's  subsidence  process,  when  it  will 
be  possible  to  detect  as  little  as  I  part  of  potato  starch 
in  a  thousand  of  wheat  flour. 

The  process  just  alluded  to  depends  on  the  fact  that 
potato  starch  has  a  higher  specific  gravity  than  wheat 
starch,  and  will  consequently  sink  in  water  sooner  than 
the  latter.  100  grams  of  the  flour  are  made  into  a  dough 
with  water,  and  the  gluten  is  separated  by  kneading 
in  water ;  the  wash  water  is  collected,  stirred  and 
passed  through  a  sieve  to  separate  the  coarser  sus- 
pended matter,  and  then  allowed  to  stand  in  a  conical 
flask  until  a  deposit  has  formed.  The  supernatant 
liquor  is  decanted  while  still  turbid,  and  the  deposit 
stirred  up  with  more  water  and  allowed  to  stand  for  a 
short  time.  After  decanting  the  turbid  supernatant 
liquor,  the  process  is  repeated  once  more,  when  the  lowest 
portion  of  the  final  deposit  will  consist  entirely  of  potato 
starch,  if  this  be  present.  The  influence  of  dilute 
potassium  hydroxide  solution  on  the  starch  grains  may 
then  be  studied  by  means  of  the  microscope,  as  directed 


268      Industrial  Organic  Analysis 

above.  Comparison  should  be  made  with  a  genuine 
sample  of  unadulterated  wheat. 

Leguminous  Starches.— Leguminous  starches  may  be 
detected  with  comparative  ease  in  wheat  flour  by  micro- 
scopic examination.  If  a  thin  layer  of  the  flour  be 
mounted  in  a  mixture  of  water  and  glycerol,  in  the 
proportion  of  2  parts  of  the  former  to  I  of  the  latter, 
and  examined  under  the  microscope  between  crossed 
Nicol  prisms,  a  selenite  plate  being  interposed  between 
the  object  and  the  lower  Nicol  prism,  the  leguminous 
starches  will  show  no  play  of  colours,  and  will  easily  be 
detected  among  the  iridescent  starch  granules  of  wheat. 
Further,  if  the  flour  is  treated  under  the  microscope  with 
a  10  to  12  per  cent,  solution  of  potassium  hydroxide,  it 
will  be  possible  to  dissolve  the  starch  of  the  leguminosae, 
leaving  a  characteristic  reticular  tissue,  made  up,  for  the 
most  part,  of  irregular  hexagons. 

Separation  of  Legumin. — Leguminous  starches  may  be 
detected  by  the  separation  of  legumin,  a  constituent 
peculiar  to  this  group.  Lecanu's  process  is  as  follows  : 
The  gluten  is  separated  in  the  usual  manner,  as  described 
above  under  the  gluten  test,  the  wash  water,  which 
contains  the  starch,  soluble  matter  and  legumin,  being 
collected,  passed  through  a  sieve  to  separate  the  coarser 
suspended  matter,  diluted,  if  necessary,  and  allowed  to 
settle.  The  supernatant  liquid  is  divided  into  2  parts  ; 
one  of  these  is  allowed  to  putrefy  or  ferment  spon- 
taneously ;  with  pure  wheat  flour,  lactic  acid  fermenta- 
tion is  the  most  common  ;  with  flours  containing 
legumin,  putrefactive  fermentation  will  set  in  at  once. 
In  lactic  acid  fermentations,  the  principal  change  which 
occurs  is  the  transformation  of  sugar  into  lactic  acid,  a 
change  which  may  be  followed  by  noting  the  increasing 
acidity  of  the  liquid  ;  in  putrefactive  fermentations,  the 


Starch  and  its  Products         269 

proteins  are  decomposed.      (See  the  account  of  the  de- 
composition of  milk  by  micro-organisms  in  Chapter  VI.) 

The  second  portion  of  the  aqueous  extract  from  the 
flour  is  filtered  clear,  and  concentrated  by  evaporation 
until  a  yellowish  scum  forms  on  the  surface  ;  it  is  then 
allowed  to  cool,  and  filtered  in  order  to  remove  the 
albumen,  which,  at  this  stage,  will  have  separated  from 
the  flour  whether  it  contains  leguminous  products  or  not. 
The  legumin,  if  present,  is  precipitated  from  the  filtrate 
by  the  addition  of  a  drop  of  acetic  acid,  filtered  off  and 
dried  in  the  steam  oven.  When  dry,  legumin  is  of  a 
horny  consistence,  insoluble  in  alcohol,  not  coloured  by 
iodine  and  easily  soluble  in  caustic  alkali  or  ammonia 
solutions,  from  which  it  may  be  precipitated  by  the 
addition  of  acid.  According  to  Lemenant  des  Chenais, 
0-9  parts  of  legumin  in  100  parts  of  flour,  represents  an 
adulteration  corresponding  to  5  per  cent,  of  added 
leguminous  flour.  Too  much  reliance  should,  however, 
not  be  placed  on  a  quantitative  estimation  of  this  kind. 

Rice  Flour. — Rice  flour  may  be  detected  in  wheat 
flour  by  a  microscopic  examination  of  the  starch  granules. 
Wheat  starch,  in  common  with  the  starches  of  rye  and 
barley,  shows  no  hilum  or  concentric  rings  in  the  majority 
of  granules,  which  are  nearly  circular,  and  flattened  in 
shape.  The  granules  of  rice  starch,  on  the  other  hand, 
are  polygonal  in  shape,  as  viewed  under  the  microscope, 
show  faint  rings,  and  under  high  magnification,  a  starred 
hilum.  Gastine  advises  staining  with  aniline  blue  or 
green  (as  previously  described),  which  will  show  up  the 
hilum  of  the  minute  starch  granules  as  a  reddish  point. 
In  all  these  microscopic  examinations,  comparison 
should  always  be  made  with  samples  which  are  known 
to  be  genuine,  as  well  as  with  samples  which  have  been 
purposely  adulterated. 


270      Industrial  Organic  Analysis 

THE  EXAMINATION  AND  ANALYSIS  OF  BARLEY. 

The  use  of  barley  for  the  preparation  of  malt  in  the 
brewing  industry  has  been  referred  to  above.  (See 
p.  248.)  The  object  of  the  tests  described  under  this 
heading  is  to  determine  the  suitability  of  barley  for 
this  purpose. 

Barley  may  vary  in  colour  from  whitish  yellow  to 
brown,  the  best  usually  being  that  of  a  bright  yellow 
colour.  If  too  light,  the  grain  may  have  been  bleached 
by  chemical  means  (sulphurous  acid)  while  if  too  dark, 
or  coloured  brown  at  the  tips,  it  has  probably  been 
unduly  exposed  to  rain.  It  should  possess  a  sweet  smell 
like  that  of  straw  ;  if  musty  or  mouldy  in  smell,  it  has 
probably  been  badly  harvested  or  stored.  The  various 
outward  characteristics  by  which  the  expert  judges  the 
quality  of  barley  will  not  be  detailed  here,  and  the 
mechanical  tests  will  only  be  briefly  referred  to.  The 
student  desiring  further  information  on  these  points,  is 
referred  to  the  works  on  the  subject  of  brewing,  mentioned 
at  the  end  of  the  present  chapter. 

Sampling. — Several  portions  are  taken  from  different 
places  and  depths  in  the  sack  or  heap,  and  well  mixed ; 
500  grams  are  then  taken  as  a  sample  for  analysis,  which 
is  preserved  in  an  air-tight  vessel. 

Mechanical  Tests. — The  more  important  mechanical 
tests  are  as  follows :  (a)  Determination  of  glumes  or  chaff, 
which  should  not  amount  to  more  than  about  7  to  loper 
cent,  of  the  grain  (Weber),  (b)  The  form  and  size  of  the 
corns  may  be  examined  by  specially  constructed  sifting 
machines  ;  (c)  the  "  Sinker  "Test ;  when  several  hundred 
corns  are  thrown  into  water,  only  about  2  to  3  per  cent, 
should  float ;  (d)  the  relative  transparency  of  the  grain 
may  be  observed  by  means  of  apparatus  specially 
designed  for  the  purpose  ;  (e)  the  weight  of  1000  corns 
(air  dried)  ;  (/)  the  weight  per  hectolitre  ;  (g)  germina- 
tion tests.  Special  apparatus  has  been  designed  for 
carrying  out  the  three  last-mentioned  tests. 


Starch  and  its  Products         271 

CHEMICAL  EXAMINATION. 

Wafer. — A  weighing  bottle,  furnished  with  a  glass 
stopper  and  measuring  about  5  to  6  cm.  in  height  and 
3-5  cm.  in  diameter,  is  carefully  dried  (in  the  air-oven) 
and  weighed  after  cooling.  5  grains  are  then  ground  up 
in  a  small  hand  mill  through  which  a  little  barley  has 
been  passed  beforehand,  and  the  meal  is  immediately 
weighed  out  in  the  closed  bottle.  After  removing  the 
stopper  the  bottle  is  placed  in  an  air-oven  at  about 
50°  to  60°,  the  temperature  being  kept  constant  for  an 
hour  and  a  half,  after  which  it  is  raised  to  105°  and  kept 
at  this  point  for  3  hours.  The  bottle  is  then  stoppered, 
cooled  and  weighed,  the  loss  in  weight  being  calculated 
to  percentage  of  water,  as  usual.  A  low  temperature  is 
employed  at  the  outset,  in  order  to  prevent  the  formation 
of  soluble  starch  and  further  decomposition  of  the  latter, 
in  the  presence  of  an  excess  of  water. 

The  water  percentage  in  barley,  which  may  vary  from 
12  to  18,  should  be  an  important  factor  in  its  commercial 
valuation  ;  a  good  barley  should  contain  from  13  to  15 
per  cent,  of  water. 

Fat. — The  dried  residue  from  the  water  determination 
is  extracted  with  ether  in  the  Soxhlet  apparatus.  (See 
p.  66.) 

Barley  may  contain  from  about  1-6  to  2-6  per  cent,  of 
fatty  matter.  Occasionally,  however,  the  barley  may 
have  been  treated  with  oil  in  order  to  give  it  a  shiny 
appearance,  in  which  case  the  fat  content  as  determined 
by  the  above  method  will  be  materially  increased. 

Nitrogen  (Proteins). — This  is  determined  on  2  to  3 
grams  of  barley,  by  the  Kjeldahl  process.  (See  p.  22.) 
The  factor  for  converting  nitrogen  into  proteins  is  6-25. 

According  to  Weber,  the  protein  content  of  barley 


272      Industrial  Organic  Analysis 

may  vary  from  6  to  18  per  cent.,  but  generally  speaking, 
a  good  barley  for  brewing  purposes  should  contain  not 
less  than  8,  and  not  more  than  n  per  cent,  of  proteins. 
A  low  protein  content  affects  the  yeast  unfavourably  in 
the  fermentation  process,  while  with  a  high  protein 
content,  the  amount  of  extract  obtained  from  the  malt  is 
decreased. 

Starch. — This  constituent  may  be  determined  on  the 
finely  ground  barley  by  the  methods  described  on  pp.  236, 
244. 

The  percentage  of  starch  in  barley  usually  lies  be- 
tween 70  and  80,  and  furnishes  a  measure  of  the  pro- 
portion of  extract  obtainable  from  the  malt  for  brewing 
purposes.  Unfortunately,  however,  the  methods  avail- 
able for  the  estimation  of  starch  in  presence  of  plant 
tissues  are  not  as  accurate  as  could  be  desired. 

Ash. — This  may  be  determined  on  the  finely  ground 
barley,  as  described  for  flour  on  p.  259. 

The  content  of  ash  in  barley  usually  lies  between 
2-5  and  3-5  per  cent. 

Test  for  Bleaching  by  Sulphur  Dioxide. — Mixed  or 
badly  coloured  barleys  may  be  made  to  assume  a  uniform 
light  yellow  colour  by  treatment  with  sulphurous  acid. 

Evidence  of  such  treatment  may  be  obtained  as 
follows  :  (Weber)  100  grams  of  the  barley  are  mixed 
with  100  c.c.  of  water,  stirred  at  intervals  for  J  an  hour, 
and  filtered.  100  c.c.  of  water  and  a  few  pieces  of  pure 
zinc  are  placed  in  a  beaker,  and  sufficient  concentrated 
hydrochloric  acid  is  added  to  cause  a  moderately  rapid 
evolution  of  hydrogen.  The  beaker  is  covered  with  a 
piece  of  filter  paper,  the  middle  of  which  has  been 
moistened  with  a  few  drops  of  lead  acetate  solution.  If, 
after  some  time,  no  brown  or  black  stain  appears  on  the 
paper,  the  zinc  and  acid  are  sufficiently  pure,  and  the 
test  may  be  proceeded  with.  The  water  is  poured  off 


Starch  and  its  Products         273 

from  the  zinc,  and  the  nitrate  from  the  barley  added  in 
its  place,  together  with  a  little  concentrated  hydrochloric 
acid.  Sulphurous  acid,  if  present,  will  be  reduced  to 
sulphuretted  hydrogen  by  the  action  of  the  nascent 
hydrogen,  and  a  brown  or  black  stain  of  lead  sulphide 
will  appear  on  the  filter  paper  covering  the  beaker.  If 
only  traces  of  sulphurous  acid  are  present,  the  stain  may 
not  appear  for  about  10  minutes. 

THE  EXAMINATION  AND  ANALYSIS  OF  MALT. 

The  method  of  preparation  and  chief  characteristics  of 
malt  have  been  outlined  above.  (See  p.  248).  As 
in  the  case  of  barley,  it  requires  considerable  experience 
to  judge  of  the  quality  of  a  malt  by  its  outward  charac- 
teristics. 

Sampling. — See  under  the  Examination  and  Analysis 
of  Barley. 

Mechanical  Tests. — Tests  similar  to  those  enumerated 
under  (b)  to  (d)  on  p.  270,  for  barley,  are  applied  also  to 
malt. 

CHEMICAL  EXAMINATION. 

Water. — This  may  be  determined  by  drying  about 
5  grams  of  the  finely  powdered  malt  in  a  weighing  bottle, 
as  described  for  barley  on  p.  271,  at  105°  for  not  more 
than  4  hours. 

Freshly  cured  malts  may  contain  from  about  i  to  3  per 
cent,  of  water,  while  malts  which  have  been  stored  for 
any  length  of  time,  may  contain  from  4  to  6  per  cent.,  or 
even  more.  Malts  containing  more  than  6  per  cent,  of 
water  may  usually  be  regarded  as  inferior  in  quality, 
yielding  beers  of  poor  taste  and  keeping  qualities.  Pale, 
or  lightly  cured  malts  generally  contain  slightly  more 
water  than  dark  malts.  Malts  which  are  known  to  have 
been  stored  for  a  long  time  and  which  contain  less  than 
3  per  cent,  of  water  should  be  regarded  with  suspicion, 
as  they  may  possibly  have  been  subjected  to  an  extra 
drying. 

I.O.A.  18 


274      Industrial  Organic  Analysis 

Preparation  of  Extract. — An  important  feature  in  the 
valuation  of  malt  for  brewing  purposes  is  the  preparation 
of  an  aqueous  extract  for  chemical  and  physical  examina- 
tion. Great  care  should  be  taken  in  carrying  out  the 
following  directions,  as  the  analytical  results  are  likely 
to  be  influenced  by  apparently  trifling  differences  in  the 
method  of  preparing  the  extract.  The  directions,  given 
by  Weber,  include  a  "  saccharification  test." 

For  the  preparation  of  the  mash,  a  500  c.c.  beaker  of 
nickel,  aluminium  or  brass  will  be  required.  The  clean, 
dry  beaker  is  first  weighed  together  with  a  thermometer, 
and  the  weight  noted.  A  little  over  50  grams  of  the 
malt  is  ground  to  a  fine  powder,  and  exactly  50  grams  of 
the  powder  weighed  out  in  the  beaker.  200  grams  of 
water  at  48°  to  50°  are  added,  and  the  beaker  is  im- 
mediately placed  in  a  water  bath  at  45°,  and  kept  at 
this  temperature  for  half  an  hour  ;  the  temperature  is 
then  raised  at  the  rate  of  i°  per  minute  up  to  70°,  and 
kept  at  this  point  for  i  hour  ;  throughout  the  whole 
mashing  process,  the  contents  of  the  beaker  should  be 
gently  stirred  with  the  thermometer.  After  the  mash 
has  been  kept  at  70°  for  10  minutes,  a  drop  is  removed 
and  tested  for  starch  by  mixing  with  a  drop  of  iodine 
solution  on  a  white  tile  ;  if  a  red  or  blue  coloration  is 
produced,  the  whole  of  the  starch  has  not  yet  been  con- 
verted into  dextrin  and  maltose,  and  the  test  is  repeated 
every  5  minutes  until  no  coloration  is  obtained  with  iodine. 

The  time  from  the  moment  at  which  the  temperature 
of  the  mash  reached  70°  to  the  complete  disappearance 
of  the  starch  is  noted  (Saccharification  Test).  With 
pale,  lightly  cured  malts,  of  high  diastatic  value,  the 
time  for  "  saccharification  "  is  usually  about  10  to  15 
minutes,  with  intermediate  malts,  15  to  20  minutes,  and 
with  dark  malts,  which  have  somewhat  lower  diastatic 
values,  owing  to  the  heavy  curing,  25  to  30  minutes. 


Starch  and  its  Products         275 

When  the  saccharification  is  complete,  the  material 
adhering  to  the  sides  of  the  beaker  is  loosened  by  means 
of  the  thermometer  and  use  of  a  wash  bottle,  and  the 
temperature  kept  at  70°  until  the  hour  is  up.  The  beaker 
is  then  removed  from  the  water  bath,  and  200  c.c.  of 
distilled  water  are  added,  or  correspondingly  less  if  much 
water  was  used  in  loosening  the  material  adhering  to  the 
sides  of  the  beaker.  The  whole  is  cooled  to  under  20°, 
the  outside  of  the  beaker  wiped  dry,  and  the  weight  of 
the  contents  made  up  to  exactly  450  grams  with  water, 
the  weight  of  the  empty  beaker  and  thermometer  having 
previously  been  determined  as  directed.  The  contents 
of  the  beaker  are  well  mixed  and  brought  on  to  a  dry 
fluted  filter  ;  the  filtrate,  or  wort,  as  it  is  called,  may  now 
be  used  for  the  determinations  of  extract,  sugar,  nitrogen, 
ash  and  phosphoric  acid,  and  depth  of  colour.  It  is  of 
interest  from  the  point  of  view  of  the  brewer  to  note  the 
odour  of  the  wort  and  whether  it  filters  clear  or  turbid. 

Determination  of  Extract. — The  specific  gravity  of  the 
wort  at  17*5°  is  determined  by  means  of  the  pyknometer. 
A  float  specially  designed  for  the  purpose  may  also  be 
used,  though  the  results  obtained  by  this  method  are  less 
accurate.  The  percentage  of  extract  by  weight  may- 
then  be  found  by  reference  to  the  accompanying  table. 

It  now  remains  to  calculate  the  percentage  of  extract 
on  the  malt  itself  (air  dried)  or  the  anhydrous  malt. 

Suppose  the  wort  to  contain  8-0  per  cent,  of  extract, 
and  the  malt  to  contain  5-0  per  cent,  of  water. 

100  parts  of  wort  will  then  contain  92  parts  of  water. 
The  weight  of  mash  before  filtering  was  450  grams, 
consisting  of  400  grams  of  water,  and  50  grams  of  malt. 
As  the  50  grams  of  malt  contained  2-5  grams  of  water, 
the  total  water  in  the  450  grams  of  mash  was  402-5 
grams.  Now  in  the  water,  92  parts  of  wort  correspond 

18— 2 


276      Industrial  Organic  Analysis 

to  8  parts  of  extract,  so  in  the  mash,  402-5  parts  of  water 
will  correspond  to 

8'°9*04°2'5  =  35-Q  parts  of  extract. 

thus,  50  grams  of  malt  yield  35  grams  of  extract,  there- 
fore the  malt  will  yield  70-0  per  cent,  of  extract. 

As  100  parts  of  malt  consist  of  5  grams  of  water  and 
95  grams  of  anhydrous  malt,  the  anhydrous  malt  will 

yield 

100 

70  X  — —  =  73-7  per  cent,  of  extract. 
95 

BALLING'S  TABLES  FOR  MALT  EXTRACT.     (Abridged) 


Sp.  G:. 

Per  cent. 
Extract. 

Sp.  Gr. 

Per  cent.  Extract 

Sp.  Gr. 

Per  cent.  Extract. 

By 

Weight 

By 

Volume 

By 

Weight. 

By 
Volume. 

By 

Weight. 

By 

Volume. 

I  -000 

o-oo 

O-OO 

1-025 

6-25 

6-4I 

•050 

12-29 

12-90 

I  -001 

0-25 

0-25 

1-026 

6-49 

6-66 

•051 

12-52 

I3-I5 

I-OO2 
1-003 

0-50 
o-75 

0-50 

o-75 

1-027 
1-028 

6-74 
6-98 

6-92 
7-18 

•052 
•°53 

12-75 
13-00 

13-41 
13-69 

1-004 

I  -00 

I-OO 

1-029 

7-25 

7-46 

•054 

13-25 

13-97 

I-OO5 

1-25 

1-26 

1-030 

7-45 

7-67 

•055 

13-48 

14-24 

1-006 

1-50 

1-51 

1-031 

7-70 

7-94 

•056 

I3-7I 

14-48 

I-OOy 

!'75 

1-76 

1-032 

7-95 

8-20 

•057 

13-95 

H'75 

1-008 

2-OO 

2-O2 

1-033 

8-20 

8-47 

-058 

14-20 

15-02 

I-OOQ 

2-25 

2-27 

1-034 

8-45 

8-74 

•059 

I4-43 

15-28 

I-OIO 

2-50 

2-52 

1-035 

8-68 

8-98 

•060 

14-67 

15-55 

I-OII 

2-75 

2-78 

1-036 

8-93 

9-25 

•061 

14-90 

15-81 

I-OI2 

3-00 

3-04 

1-037 

9-17 

9-47 

•062 

I5-I5 

16-09 

I-OI3 

3-25 

3'29 

1-038 

9-41 

9-77 

•063 

15-37 

16-34 

I-OI4 

3-50 

3'55 

1-039 

9-65 

10-03 

•064 

15-60 

1  6-60 

I-OI5 

3-75 

3-81 

1-040 

9-90 

10-31 

•065 

15-84 

16-87 

I-OI6 

4-00 

4-06 

1-041 

10-15 

10-57 

•066 

16-07 

J7-i3 

I-Oly 

4-25 

4-32 

1-042 

10-38 

10-81 

•067 

16-30 

17-39 

1-018 

4-50 

4-58 

1-043 

10-61 

11-07 

•068 

16-53 

17-65 

1-019 

4'75 

4-84 

1-044 

10-85 

n-33 

•0.69 

16-77 

17-93 

i  -020 

5-00 

5-10 

1-045 

II-IO 

1  1  -60 

•070 

17-00 

18-19 

I-O2I 

5-25 

5-36 

1-046 

n-33 

11-87 

•071 

17-22 

18-44 

I-O22 

5-50 

5-62 

1-047 

II-59 

12-13 

•072 

I7-45 

18-71 

1-023 

5-75 

5-88 

1-048 

n-8o 

12-37 

•073 

17-67 

18-95 

1-024 

6-00 

6-14 

1-049 

12-05 

12-64 

•074 

17-90 

-9-23 

Starch  and  its  Products         277 

The  yield  of  extract  on  the  anhydrous  malt  usually 
lies  between  72  and  80  per  cent. 

It  is  of  interest  to  the  brewer  to  know  the  yield  of 
extract  obtainable  from  malt,  as  this,  together  with  the 
added  brewing  sugar1  (invert  sugar  or  glucose)  con- 
stitutes the  fermentable  material  used  in  the  brewing  of 
beer.  The  yield  is  usually  stated  in  Ibs.  per  quarter 
(281bs.). 

Sugar. — A  portion  of  the  wort,  obtained  as  described 
above,  is  diluted  to  10  times  its  volume,  and  the  sugar 
determined  as  described  on  p.  140.  (See  also  p.  242.) 
The  copper  oxide  obtained  may  be  converted  to  maltose 
by  reference  to  the  table  on  p.  243. 

In  order  to  find  the  percentage  of  sugar  in  the  extract, 
the  percentage  of  extract  by  volume  in  the  wort  is  found 
by  reference  to  the  table  on  p.  276.  Having  determined 
the  sugar  in  a  given  volume  of  wort,  the  sugar  in  the 
extract  corresponding  to  this  volume  of  wort  is  easily 
calculated,  and  hence  the  percentage  of  sugar  in  the 
extract. 

Subtracting  the  percentage  of  sugar  in  the  extract 
from  100,  the  percentage  of  matter  other  than  sugar  in 
the  extract  is  found. 

According  to  Weber,  the  ratio  of  the  sugar  to  extract 
less  sugar,  varies  from  i  :  0-4  or  0-5  in  pale  (Pilsner)  malts, 
to  i  :  0-6  or  0-7  in  dark  (Munich)  malts. 

Nitrogen. — 25  c.c.  of  the  wort,  obtained  as  described 
above,  are  evaporated  to  a  syrup,  and  the  nitrogen,  and 
hence  the  proteins,  determined  in  the  latter  by  the 
Kjeldahl  process,  as  described  on  p.  22.  As  in  the  case 
of  malt,  the  factor  for  converting  nitrogen  into  proteins 
is  6-25. 

The  percentage  of  nitrogen,  cakulated  on  the  extract, 
varies  from  0-5  to  0-8,  and  the  proteins  from  3  to  5. 

i  Brewer's  "glucose"  contains,  besides  glucose,  maltose  and  dextrin. 


278      Industrial  Organic  Analysis 

The  nitrogen  may  also  be  determined  direct,  on  the 
finely  ground  malt. 

Although  both  in  barley  and  malt  the  whole  of  the 
nitrogen  is  not  present  as  proteins,  the  determination  of 
nitrogen  is  nevertheless  of  value  when  used  comparatively 
especially  in  its  bearing  on  the  physiology  of  the  fer- 
mentation process. 

Ash. — 50  c.c.  of  the  wort  are  evaporated  to  dryness  in 
a  weighed  platinum  dish,  and  the  residue  incinerated. 
(See  p.  259.) 

The  content  of  ash  in  the  extract  (refer  to  tables  as 
explained  above  for  sugar)  varies  from  1-4  to  1-9  per 
cent. 

Acidity  of  Malt. — 100  grams  of  finely  powdered  malt 
are  well  mixed  with  400  c.c.  of  20  per  cent,  alcohol,  and 
allowed  to  stand  for  6  hours,  the  mixture  being  shaken 
from  time  to  time.  The  mixture  is  then  filtered,  and 
100  *c.c.  of  the  filtrate  are  titrated  with  decinormal 
sodium  hydroxide  or  baryta  solution,  until  a  drop 
brought  on  to  a  piece  of  sensitive  blue  litmus  paper  no 
longer  produces  a  red  coloration.  The  number  of  cubic 
centimetres  of  decinormal  alkali  used  is  calculated  to 
lactic  acid  by  multiplying  it  by  0-009. 

The  percentage  of  acid,  calculated  as  lactic  acid,  in 
malt,  varies  from  0-15  to  0-4. 

Diastatic  Value.— The  determination  of  the  diastatic 
value  of  malt  and  malt  extract  is  described  on  p.  249. 

Colour  of  the  Wort. — The  depth  of  colour  of  the  wort, 
obtained  as  described  above,  is  determined  by  means  of 
specially  designed  colorimeters,  comparison  being  made 
with  the  colour  of  iodine  solutions  of  known  strength. 

REFERENCES. 

"  Foods,  their  Composition  and  Analysis,"  by  A.  and  M. 
Wynter  Blyth.  1909,  Chsfe.  Griffin  &  Co.,  London. 

Allen's  "Commercial  Organic  Analysis"  (revised  edition), 
Vols.  I.  and  VIII.  1909,  J.  &  A.  Churchill,  London. 


Starch  and  its  Products         279 

"  Chemisch-Technische  Untersuchungsmethoden,"  Vol.  IV., 
by  Lunge.  1911,  J.  Springer,  Berlin. 

"  Chemisch-Technische  Analyse,"  Vol.  II.,  by  Post.  1908, 
Fr.  Vieweg,  Brunswick. 

"  Food  Inspection  and  Analysis,"  by  Leach.  1909,  Wiley 
&  Sons,  New  York. 

"  Anleitung  zu  den  Chemisch-Technischen  Untersuchungs- 
methoden fiir  Brauer,"  by  F.  Weber.  1904,  J.  L.  Stich,  Nurn- 
berg. 

"  Select  Methods  of  Food  Analysis,"  by  Leffmann  &  Beam. 
1905,  Rebman,  Ltd.,  London. 

"  Laboratory  Studies  for  Brewing  Students,"  by  A.  J.  Brown. 
1904,  Longmans,  Green  &  Co.,  London. 

"  The  Brewer's  Analyst,"  by  Douglas  Bailey.  1907,  Kegan, 
Paul,  Trubner  &  Trench,  London. 

"  The  Student's  Manual  of  Yeast  Culture, "by  Grove  Johnson. 
Published  by  the  Author,  London,  1908. 

"  Brewing,"  by  Chaston  Chapman.  1912,  Cambridge  Univer. 
sity  Press. 

"The  Principles  and  Practice  of  Brewing,"  by  Sykes  &  Ling. 
1911,  Chas.  Griffin,  London. 

"  Illustriertes  Brauereilexikon,"  edited  by  Max  Delbriick. 
Paul  Parey,  Berlin. 

Reports  to  the  Local  Government  Board  on  Public  Health  and 
Medical  Subjects,  New  Series,  No.  49,  Food  Report,  No.  12, 
1911  :  "On  the  Bleaching  of  Flour  and  the  Addition  of  So-called 
'Improvers'  to  Flour,"  by  Dr.  J.  M.  Hamill  ;  and  "On  the 
Chemical  Changes  produced  in  Flour  by  Bleaching,"  by  Dr.  G.  W. 
Monier- Williams.  Wyman  &  Sons,  London. 

Ditto,  New  Series,  No.  55,  Food  Report,  No.  14  :  Dr.  J.  M. 
Hamill's  Report  to  the  Local  Government  Board  on  the  "  Nutri- 
tive Value  of  Bread  made  from  different  Varieties  of  Wheat 
Flour." 


CHAPTER   VIII 

PRESERVATIVES  AND  ARTIFICIAL  COLOURING  MATTERS  IN 
FOODS 

PART  I. — PRESERVATIVES. 
INTRODUCTORY. 

ALL  foodstuffs  are  liable  to  decomposition  and  decay 
through  the  agency  of  micro-organisms,  and  if  they  are 
to  be  kept  for  any  length  of  time  in  a  state  fit  for  human 
consumption,  it  is  sometimes  necessary  to  adopt  some 
means  for  their  preservation,  by  which  the  micro- 
organisms which  they  contain  may  either  be  destroyed 
or  temporarily  rendered  sufficiently  inactive  to  prevent 
them  from  appreciably  affecting  the  food. 

The  method  of  heat  sterilisation  and  subsequent 
preservation  in  hermetically  closed  vessels,  as  practised 
in  the  canning  industry,  has  for  its  object  the  destruction 
of  the  micro-organisms  by  heat  (see  Chapter  VI.)  and 
the  protection  of  the  food  from  further  contamination 
until  it  reaches  the  consumer.  Preservation  by  cold 
storage  is  based  on  the  fact  that  micro-organisms  become 
inactive  at  low  temperatures,  while  the  treatment  of  the 
food  with  wood  smoke,  salt  or  other  chemicals  has  a 
similar  effect,  i.e.,  the  temporary  checking  of  the  activity 
of  the  micro-organisms,  without,  however,  destroying 
them. 

Heat  sterilisation,  if  efficiently  carried  out,  is  naturally 
more  thorough  in  its  effects  than  the  other  methods  just 
mentioned,  though  it  has  a  real  disadvantage  when 
applied  to  many  foodstuffs  owing  to  the  partial  decom- 
position suffered  by  proteins  and  carbohydrates  under 
the  influence  of  heat.  (See  Chapter  VI.,  the  Heat 
Sterilisation  and  Pasteurising  of  Milk,  also  Condensed 
Milk.) 


Preservatives,  etc.,  in  Foods     281 

In  cases  when,  for  this  and  other  reasons,  the  food 
cannot  be  completely  sterilised  by  heat,  recourse  may 
also  be  had  to  cold  storage  and  sometimes  also  the 
addition  of  chemical  preservatives.  Cold  storage  and 
addition  of  preservatives  may  also  be  practised  simul- 
taneously. The  merits  and  disadvantages  of  the 
latter  method  will  be  discussed  below  ;  it  may,  however, 
be  pointed  out  here  that,  generally  speaking,  it  would 
probably  be  in  the  best  interest  of  the  consumer  if  the 
method  of  cold  storage  were  employed  in  preference  to 
the  addition  of  preservatives,  wherever  possible. 

In  the  present  chapter  we  are  concerned  with  the 
method  of  preserving  by  the  addition  of  certain  chemicals 
with  the  object  of  inhibiting  the  growth  and  development 
of  micro-organisms  in  the  food  ;  among  the  substances 
which  have  been  used  in  this  capacity,  the  following  may 
be  mentioned  :  Common  salt,  boric  acid  and  borax, 
sodium  fluoride,  sulphites  and  sulphurous  acid,  benzoic 
acid  and  benzoates,  salicylic  acid  and  salicylates, 
formaldehyde,  /3  naphthol,  abrastol  or  asaprol  (the 
calcium  salt  of  /3  naphthol  sulphonic  acid),  hydrogen 
peroxide,  saccharin,  formic  acid  and  formates.  As 
regards  the  advisability  of  permitting  such  substances 
to  be  added  to  foods,  two  main  questions  come  into 
consideration  : — 

Firstly,  the  question  as  to  whether  the  substance 
added  is  injurious  to  health,  even  in  the  minutest  pro- 
portions, or  has  an  undesirable  effect  on  the  food  with 
which  it  comes  into  contact,  and  if  not,  the  maximum 
proportion  in  which  it  can  be  added  without  prejudice 
to  the  purchaser.  While  the  use  of  such  a  preservative 
as  common  salt  has  long  been  recognised  and  sanctioned 
in  all  countries,  considerable  diversity  of  opinion  still 
exists  as  to  the  advisability  of  permitting  the  use  of 
certain  preservatives,  such  as  borax,  boric  acid  or 
benzoic  acid  in  small  quantities  ;  other  preservatives, 
such  as  sodium  fluoride  and  formaldehyde,  are  generally 
looked  on  as  distinctly  injurious  to  health. 

Secondly,  there  is  the  question  as  to  whether,  with 
proper  organisation,  care  and  cleanliness,  it  is  possible 
to  supply  a  given  article  of  food  to  the  consumer  in  a 


282      Industrial  Organic  Analysis 

fresh  and  pure  state,  without  the  addition  of  preserva- 
tives. If  this  be  the  case,  then  the  use  of  preservatives 
must  obviously  be  looked  on  as  decidedly  objectionable, 
as  it  may  often  be  practised  in  order  to  temporarily 
mask  uncleanliness  in  treatment.  This  point  will  be 
further  discussed  below,  in  dealing  with  the  subject  of 
preservatives  in  milk. 

The  law  of  the  United  Kingdom  does  not  definitely 
prohibit  the  use  of  any  given  substance  as  a  preservative 
for  foods,  nor  does  it  actually  set  any  definite  limit  as  to 
the  proportion  in  which  any  given  preservative  may  be 
used. 

Section  3  of  the  Sale  of  Foods  and  Drugs  Act  of  1875 
provides  that  "  No  person  shall  mix,  colour,  stain,  or 
powder  ...  any  article  of  food  with  any  ingredient  or 
material  so  as  to  render  the  article  injurious  to  health, 
.  .  .  under  a  penalty  in  each  case  not  exceeding  fifty 
pounds  for  the  first  offence  ;  every  offence,  after  convic- 
tion for  a  first  offence,  shall  be  a  misdemeanour,  for 
whicji  the  person,  on  conviction,  shall  be  imprisoned  for 
a  period  not  exceeding  six  months  with  hard  labour." 

Very  few  prosecutions  have  been  instituted  under  this 
section,  and  practically  all  prosecutions  are  instituted 
under  section  6  of  the  same  Act,  which  is  to  the  following 
effect  :— 

"  No  person  shall  sell  to  the  prejudice  of  the  purchaser 
any  article  of  food  or  any  drug  which  is  not  of  the  nature, 
substance,  and  quality  of  the  article  demanded  by  such 
purchaser,  under  a  penalty  of  twenty  pounds.  ..." 
Exceptions  are  made  in  the  case  of  foreign  matter  not 
injurious  to  health,  necessarily  or  unavoidably  added  in 
the  preparation  of  the  food  or  drug,  and  not  with  the 
object  of  fraudulently  increasing  its  bulk  or  weight,  or 
concealing  its  inferior  quality  ;  also  in  the  case  of  pro- 
prietary or  patent  medicines,  supplied  in  a  state  required 
by  the  specification  of  the  patent. 

The  purchaser  is  not  held  to  be  prejudiced  if  notice 
was  given  him  at  the  time  of  the  sale  that  the  article 
sold  was  not  of  the  nature,  substance  and  quality  of  the 
article  demanded  ;  further,  in  order  to  show  that  the 
article  was  sold  to  the  prejudice  of  the  purchaser,  it  is 


Preservatives,  etc.,  in  Foods     283 

not  necessary  to  show  that  he  has  sustained  actual 
prejudice  or  damage. 

Questions  as  to  whether  the  purchaser  shall  be  held 
to  have  been  prejudiced  by  the  addition  of  material 
injurious  to  health  to  the  article  supplied,  or  otherwise, 
are  usually  decided  with  reference  to  the  opinions  and 
recommendations  of  the  Medical  Officers  of  Health  and 
other  experts.  With  the  exception  of  the  Public  Health 
(Milk  and  Cream)  Regulations  of  1912,  referred  to  below 
under  the  heading  of  Preservatives  in  Milk,  Cream, 
Butter  and  Margarine,  no  further  legislation  has  resulted 
from  such  recommendations.  This  matter  will  be  more 
fully  dealt  with  below,  under  the  various  headings. 

In  the  United  States  the  law  forbids  the  sale  of  food 
containing  poisonous  or  deleterious  substances  which 
may  render  the  food  injurious  to  health,  but  does  not 
definitely  prohibit  or  limit  the  use  of  any  given  preserva- 
tive. Some  preservatives,  such  as  boric  and  salicylic 
acids,  and  formaldehyde,  are,  however,  held  to  be 
injurious  to  health,  and  prosecutions  have  been  success- 
fully maintained  against  them  ;  the  matter  is  at  present 
under  investigation. 

In  France,  Germany,  Austria-Hungary  and  Holland, 
the  law  is,  generally  speaking,  more  explicit  and  stringent 
with  regard  to  the  use  of  preservatives  ;  either  certain 
preservatives  may  be  totally  prohibited,  or  the  use  of 
preservatives  in  certain  foods  may  be  forbidden. 

PRESERVATIVES   IN   MILK,    CREAM,    BUTTER   AND 
MARGARINE. 

The  provisions  of  the  Public  Health  (Milk  and  Cream) 
Regulations  of  1912  are,  briefly,  as  follows  :  The  addition 
of  preservative  substance  of  any  kind  to  milk  intended 
for  sale  for  human  consumption  is  absolutely  prohibited  ; 
also  the  sale,  or  offering,  exposing  or  keeping  for  sale,  of 
milk  containing  any  preservative. 

The  addition  of  preservatives  to  cream  which  contains 
less  than  35  per  cent,  of  fat  is  prohibited,  and  the  only 
preservatives  which  may  be  added  to  cream  which  con- 
tains more  than  35  per  cent,  of  fat  are  boric  acid  and 


284      Industrial  Organic  Analysis 

borax,  or  mixtures  of  these,  and  hydrogen  peroxide. 
Cream  containing  preservative  must  be  described  as 
"  Preserved  Cream,"  on  a  label  on  the  vessel  in  which  it 
is  sold,  and  in  all  advertisements,  price  lists,  etc.,  used  in 
connection  with  its  sale  ;  if  it  contains  boric  acid  or 
borax,  the  amount  of  these  substances,  calculated  as 
boric  acid  (H3BO8)  must  be  accurately  stated  on  the 
label  as  not  exceeding  a  certain  limit. 

The  main  reasons  why  preservatives  should  be 
excluded  from  milk  are  stated  in  the  circular  letter  of  the 
Local  Government  Board,  dated  July  nth,  1906,  as 
follows  :  "  Under  the  influence  of  these  preservatives,1 
milk  may  be  exposed  without  sensible  injury  to  con- 
ditions which  would  otherwise  render  it  unsaleable.  It 
may  remain  sweet  to  taste  and  smell,  and  yet  have 
incorporated  disease  germs  of  various  kinds,  whereof  the 
activity  may  be  susperided  for  a  time  by  the  action  of  the 
preservative,  but  may  be  resumed  again  after  the  milk 
has  been  digested.  The  Committee,  after  hearing  evi- 
dence, from  milk  traders,  concluded  that  the  addition  of 
preservatives  to  milk  is  not  necessary  for  the  purposes  of 
the  milk  trade  .  .  .  and  the  Committee  recommended 
that  no  preservatives  should  be  added  to  milk."  A 
further  important  reason  for  the  above  recommendation 
was  the  fact  that  milk  is  largely  consumed  by  children 
and  invalids,  i.e.,  individuals  who  would  be  especially 
susceptible  to  the  harmful  influence  of  any  preservative. 

The  addition  of  preservatives  to  milk  is  prohibited  in 
most  other  countries. 

With  regard  to  cream,  no  definite  limit  has  been  set 
on  the  amount  of  boric  acid  which  may  be  used  for  its 
preservation.  In  his  report  to  the  Local  Government 
Board  on  the  Use  of  Preservatives  in  Cream,  1909, 
Dr.  J.  M.  Hamill  recommended  that  the  maximum 
amount  of  boric  acid  (H3BO3)  should  ba  0-4  per  cent, 
from  May  to  October,  inclusive,  and  0-25  per  cent. during 
the  rest  of  the  year  ;  for  the  present,  however,  it  is 
probable  that  prosecution  under  the  Sale  of  Foods  and 
Drugs  Act  would  not  succeed  for  less  than  0-5  per  cent. 

1  The  evidence  on  which  these  observations  were  based  was  chiefly  in  connection 
with  boric  acid  and  formaldehyde. 


Preservatives,  etc.,  in  Foods     285 

of  boric  acid.  In  the  report  mentioned  above,  it  is 
pointed  out  that  as  cream  is  used  to  a  considerable  extent 
as  food  for  children  and  invalids,  individuals  specially 
sensitive  to  boric  acid,  the  declaration  of  the  presence 
of  this  substance  should  be  made  obligatory,  in  order  that 
it  may  be  avoided  by  those  who  object  to  it. 

In  France  and  most  of  the  United  States  of  America, 
the  use  of  preservatives  in  cream  is  prohibited  by  law  ; 
in  Germany,  persons  selling  preserved  cream  are  liable 
to  proceedings  under  the  Nahrungsmittelgesetz  of  1879. 

As  regards  butter  and  margarine,  it  was  recommended 
in  the  circular  letter  of  the  Local  Government  Board 
referred  to  above,  that  the  only  preservative  allowed  in 
these  foods  should  be  boric  acid  or  borax,  in  proportions 
not  exceeding  0-5  per  cent.,  expressed  as  (H3BO3).  It  is 
probable  that  for  the  purposes  of  the  Sale  of  Foods  and 
Drugs  Act,  the  limit  would  be  placed  at  0-5  per  cent,  of 
boric  acid ;  thus,  a  prosecution  for  51  grains  per  pound l  of 
boric  acid  in  margarine  succeeded,  while  a  conviction  for 
25  grains  per  pound  of  boric  acid  in  butter  was  quashed 
at  the  Quarter  Sessions.  It  is,  however,  doubtful  whether 
the  use  of  any  preservative  other  than  boric  acid  (or 
salt)  would,  in  itself,  be  considered  an  offence. 

In  Germany,  Holland,  and  some  other  countries, 
preservatives,  except  salt,  are  prohibited  in  butter  and 
margarine. 

Before  starting  on  the  description  of  the  methods  for 
the  detection  of  preservatives  in  milk,  cream,  butter  and 
margarine,  it  may  be  mentioned  that  milk  is  rarely 
treated  with  preservatives  ;  on  account  of  the  importance 
of  excluding  preservatives  from  milk,  however,  methods 
for  detecting  most  of  the  commoner  preservatives  in  this 
article  of  food  are  given  below. 

Sodium  Carbonate  or  Bicarbonate  in  Milk. — These  sub- 
stances have  no  antiseptic  action,  but  are  added  in  order 
to  neutralise  the  lactic  acid  produced  in  the  souring  of 
the  milk  (see  Chapter  VI.),  and  thus  to  prevent  coagula- 
tion. The  growth  of  the  bacteria  is  not  impeded. 

According  to  Fade,  the  ash  of  10  c.c.  of  genuine  milk 

i  35  grains  per  pound  =  o'5  per  cent. 


286      Industrial  Organic  Analysis 

should  require  only  i  drop  of  decinormal  acid  for 
neutralisation  (indicator  methyl  orange)  ;  if  an  alkali 
carbonate  has  been  added,  more  acid  will  be  required  for 
neutralisation  of  the  ash. 

The  estimation  of  added  carbonate  by  this  method 
may  be  interfered  with  owing  to  the  conversion  of  some 
of  the  carbonate  to  phosphate  during  the  incineration  ; 
in  order  to  allow  for  this,  it  is  recommended  that  the 
soluble  phosphate  should  be  determined  in  the  ash, 
recalculated  to  sodium  carbonate,  and  added  to  the 
amount  of  carbonate  already  found  by  titration. 

Soxhlet  and  Scheibe  estimate  the  carbon  dioxide  in  the 
ash  of  the  milk  ;  they  state  that  ash  from  genuine  milk 
should  not  yield  over  2  per  cent,  of  carbon  dioxide. 

Sodium  carbonate  has  been  known  to  have  been  added 
to  milk  in  the  proportion  of  I  part  of  the  anhydrous  salt 
per  litre. 

Detection  of  Boric  Acid  (or  Borax),  (a)  In  Milk  or 
Cream. — 10  c.c.  of  the  milk  or  cream  are  made  alkaline 
with  milk  of  lime,  evaporated  to  dryness,  and  the  residue 
ignited  until  all  organic  matter  is  destroyed  and  con- 
verted into  carbon.  The  ash  is  dissolved  in  the  smallest 
possible  quantity  of  dilute  hydrochloric  acid,  and  the 
solution  filtered  from  the  carbonaceous  residue.  The 
filtrate  is  evaporated  to  dryness  in  order  to  expel  com- 
pletely the  excess  of  acid  ;  for  this  purpose,  the  residue 
may,  if  necessary,  be  treated  with  a  further  quantity  of 
water,  and  the  solution  again  evaporated  to  dryness. 
The  residue  is  then  moistened  with  a  little  very  dilute 
hydrochloric  acid,  and  a  little  tincture  .of  turmeric  is 
stirred  into  the  crystalline  mass.  If  the  least  trace  of 
boric  acid  be  present,  a  cherry-red  coloration  will  result. 
In  this  way  the  presence  of  as  little  as  o-ooi  to  0-002  per 
cent,  of  boric  acid  in  milk  or  cream  may  be  shown.  It 


Preservatives,  etc.,  in  Foods     287 

should  be  noted  that  concentrated  hydrochloric  acid  also 
produces  a  cherry-red  coloration  with  turmeric,  but  in  the 
absence  of  boric  acid,  this  coloration  would  be  discharged 
on  dilution. 

If  preferred,  the  ash  may  also  be  tested  for  boric  acid 
by  the  well-known  green  flame  reaction,  as  described 
below  on  p.  298. 

(b)  In  Butter  or  Margarine. — Boric  acid  may  be 
detected  by  examining,  as  described  above,  the  aqueous 
serum  which  collects  beneath  the  fat  when  the  butter 
or  margarine  is  melted  and  allowed  to  stand  in  a  cylindri- 
cal vessel  at  about  60°  for  an  hour  or  two.  Boric  acid 
may  also  be  tested  for  in  the  water  soluble  portion  of  the 
non-fatty  solids,  obtained  as  described  in  Chapter  VI., 
p.  230. 

The  Estimation  of  Boric  Acid,  (a)  In  Milk. — A  known 
quantity  of  milk,  say  100  grams,  is  made  alkaline  with 
milk  of  lime,  evaporated  to  dryness  and  incinerated  as 
described  under  the  determination  of  ash  in  milk  in 
Chapter  VI.,  p.  208. 

The  residue  is  dissolved  in  a  convenient  quantity  of 
water  to  which  a  little  hydrochloric  acid  has  been  added, 
and  the  boric  acid  determined  in  the  solution  as  described 
below,  under  the  next  heading. 

(b}  In  Butter,  Margarine  and  Cream. — 50  grams  of  the 
sample  are  warmed  and  well  shaken  with  50  c.c.  of  water. 
After  cooling  and  allowing  the  layers  to  separate,  most 
of  the  aqueous  layer  is  syphoned  off,  50  c.c.  thereof 
placed  in  a  100  c.c.  measuring  flask  and  heated  in  boiling 
water,  while  decinormal  sulphuric  acid  is  added  in  small 
quantities  from  time  to  time,  until  the  casein  separates 
as  a  flocculent  precipitate.  The  solution  is  then  cooled, 
made  up  to  100  c.c.  and  filtered  through  a  dry  filter  ; 
50  c.c.  of  the  filtrate  may  then  be  measured  off  for  the 


288      Industrial  Organic  Analysis 

boric  acid  titration,  the  rest  being  preserved  for  the  salt 
estimation.  The  method  adopted  for  the  estimation  of 
the  boric  acid  depends  on  the  fact  that  this  substance 
may  be  titrated  with  sodium  hydroxide  solution  and 
phenol  phthalein  as  indicator,  in  presence  of  glycerol. 
The  solution  for  the  titration  is  first  made  neutral  to 
methyl  orange  by  adding  a  slight  excess  of  deci- 
normal  sodium  hydroxide  solution,  and  then  decinormal 
sulphuric  acid,  till  a  pink  tint  just  appears  ;  glycerol 
previously  made  neutral  to  phenol  phthalein  with 
sodium  hydroxide  is  then  added  in  sufficient  amount  to 
make  up  about  one-third  of  the  bulk  of  the  liquid,  which 
is  then  titrated  with  decinormal  sodium  hydroxide 
solution  in  presence  of  a  few  drops  of  phenol  phthalein 
solution,  till  a  pink  tint  appears.  I  c.c.  of  decinormal 
alkali  corresponds  to  0-0062  gram  of  boric  acid  (H3BO8), 
I  molecule  of  boric  acid  neutralising  I  molecule  of  sodium 
hydroxide,  under  the  conditions  of  the  titration. 

In  calculating  the  amount  of  boric  acid  in  the  sample, 
the  amount  of  water  in  the  latter  must  be  allowed  for. 
Thus,  if  the  butter  contained  14  per  cent,  of  water,  the 
total  volume  of  the  aqueous  layer  would  be  57  c.c.,  of 
which  50  c.c.  were  diluted  to  100  c.c.,  of  which,  again, 
50  c.c.  were  taken  for  the  titration.  The  estimation  of 
water  in  butter  and  margarine  has  already  been  described 
in  Chapter  III.,  p.  85.  The  amount  of  water  in  cream 
may  be  arrived  at  with  sufficient  accuracy  for  the  present 
purpose,  by  subtracting  the  percentage  of  fat  from  100. 
The  estimation  of  fat  in  cream  is  described  in  Chapter  VI., 
p.  197. 

The  Estimation  of  Salt  in  Butter  and  Margarine. — 
10  c.c.  of  the  filtered  slightly  acid  liquid  as  used  for  the 
boric  acid  titration  are  diluted  with  distilled  water  to 
about  50  c.c.,  and  titrated  with  decinormal  silver  nitrate 


Preservatives,  etc.,  in  Foods     289 

solution,  with  a  few  drops  of  potassium  chromate  as 
indicator,  in  the  usual  way.  The  calculation  of  the 
percentage  of  salt  in  the  sample  is  made  on  the  same 
lines  as  indicated  above  for  boric  acid. 

Detection  of  Salicylic  and  Benzoic  Acids. — These  sub- 
stances may  be  added  as  such,  or  in  the  form  of  their 
sodium  salts.  In  most  countries  their  use  as  preserva- 
tives for  butter  or  margarine  is  not  allowed,  and  in 
France  they  are  definitely  prohibited  by  law,  in  all  foods. 

(a)  In  Milk. — Girard  recommends  the  following  pro- 
cess for  the  detection  of  salicylic  acid  in  milk  ;  it  may 
also  be  applied  for  the  detection  of  benzoic  acid  :  100  c.c. 
of  the  milk  are  diluted  with  100  c.c.  of  water  at  60°, 
treated  with  8  drops  of  acetic  acid  and  8  drops  of  a 
saturated  solution  of  mercuric  nitrate,  shaken  and 
filtered.  The  acid  filtrate,  which  has  thus  been  freed 
from  fat  and  proteins,  is  extracted  with  ether,  which  will 
take  up  the  salicylic  or  benzoic  acid.  The  ethereal  layer 
is  separated  off,  filtered  through  a  dry  filter  and  allowed 
to  evaporate  spontaneously  in  a  dish  ;  salicylic  or 
benzoic  acid,  if  present,  will  then  be  obtained  as  a  white 
crystalline  powder.  For  identification,  the  crystals  may 
be  sublimed  on  to  a  watch  glass  and  tested  fcr  their 
melting  point ;  salicylic  acid  melts  at  155°  to  156°,  and 
benzoic  acid  at  121°.  The  following  reactions  may  also 
be  used  for  the  identification  of  these  acids  :  In  neutral 
aqueous  or  alcoholic  solution,  salicylic  acid  gives  a  fine 
violet  coloration  on  the  addition  of  a  drop  or  two  of 
neutral  ferric  chloride  solution,  and  benzoic  acid  gives  a 
buff-coloured  precipitate  with  the  same  reagent  in  neutral 
aqueous  solution.  If  an  aqueous  solution  of  benzoic 
acid  is  warmed  for  5  to  10  minutes  on  the  water  bath 
with  a  few  cubic  centimetres  of  a  0-5  per  cent,  solution  of 
hydrogen  peroxide,  the  benzoic  acid  is  partially  con- 

I.O.A.  IQ 


290      Industrial  Organic  Analysis 

verted  into  salicylic  acid,  which  may  be  recognised  by  the 
reaction  just  described.  A  delicate  reaction  for  the 
detection  of  benzoic  acid  in  the  absence  of  salicylic  acid 
is  given  under  the  next  heading.  (See  also  p.  306.) 

(b)  In  Butter  and  Margarine. — The  following  method 
has  been  devised  by  Robin  for  the  detection  of  benzoic 
acid  ;  it  may  also  be  adapted  for  the  detection  of 
salicylic  acid  :  25  grams  of  the  melted  butter  or  mar- 
garine are  shaken  in  a  separating  funnel  with  50  c.c.  of 
a  i  per  cent,  solution  of  sodium  bicarbonate,  15  c.c.  of 
alcohol  being  added  in  order  to  facilitate  the  separation 
of  the  fat  from  the  aqueous  liquid.  After  shaking  with 
a  rotary  motion,  so  as  to  avoid  the  formation  of  a 
troublesome  emulsion,  and  allowing  the  layers  to 
separate,  the  aqueous  alcoholic  liquid  is  run  off,  treated 
with  7  to  8  drops  of  sulphuric  acid,  heated  to  boiling, 
shaken  with  a  little  fuller's  earth  (in  order  to  facilitate 
the  separation  of  the  proteins),  and  filtered  through  a  wet 
filter.  The  cooled  filtrate  is  shaken  out  with  40  c.c.  of 
ether,  and  the  ethereal  extract,  containing  the  benzoic 
or  salicylic  acid,  is  separated  off  and  shaken  with  20  c.c. 
of  water  and  5  c.c.  of  alcohol  to  remove  excess  of  mineral 
acid,  and  then  with  20  c.c.  of  a  I  per  cent,  solution  of 
sodium  bicarbonate  and  5  c.c.  of  alcohol.  The  alkaline 
solution,  which  will  contain  the  benzoic  or  salicylic  acids 
as  sodium  salts,  is  run  into  a  small  dish  and  evaporated 
to  dryness  on  the  water  bath.  A  small  portion  of  the 
residue  may  now  be  tested  for  salicylic  acid  by  dissolving 
in  water  and  adding  neutral  ferric  chloride  to  the  neutral 
solution  ;  if  salicylic  acid  be  absent,  the  bulk  of  the 
residue  may  be  tested  for  benzoic  acid  by  the  following 
delicate  reaction  :  Add  5  c.c.  of  concentrated  sulphuric 
acid  and  10  drops  of  fuming  nitric  acid,  and  heat  carefully 
until  white  fumes  are  given  off.  Pour  the  resulting 


Preservatives,  etc.,  in  Foods     291 

mixture,  which  should  be  colourless  or  light  yellow,  into 
50  c.c.  of  water,  and  add  sufficient  concentrated  ammonia 
solution  to  give  a  decidedly  alkaline  reaction.  After 
cooling,  add  ammonium  sulphide  solution  drop  by  drop  ; 
if  benzoic  acid  was  present  in  the  sample,  an  orange 
yellow  coloration  will  be  produced,  depending  on  the 
formation  of  the  ammonium  salt  of  an  amido  nitro 
benzoic  acid.  By  this  method,  the  presence  of  0-05  per 
cent,  of  benzoic  acid  in  the  sample  may  easily  be  detected. 
(See  also  p.  306.) 

The  Detection  of  Formaldehyde  in  Milk.— This  pre- 
servative is  generally  looked  on  as  poisonous  ;  a  further 
objection  to  its  use  is  that  it  enters  into  chemical  com- 
bination with  proteins.  It  should  be  tested  for  in  milk 
before  the  sample  is  too  old,  as  it  (the  formaldehyde) 
disappears  in  time. 

The  best  method  for  the  detection  of  formaldehyde  in 
milk  is  to  add  i  c.c.  of  dilute  sulphuric  acid  (i  part  acid 
to  3  parts  water)  to  100  c.c.  of  the  sample  and  to  distil 
until  the  distillate  measures  20  c.c.  10  c.c.  of  the  dis- 
tillate are  tested  with  Scruffs  reagent,  i.e.,  a  solution  of 
magenta  which  has  been  bleached  by  adding  just  the 
requisite  quantity  of  sulphurous  acid.  If,  after  5  to  10 
minutes,  an  intense  carmine  red  coloration  is  formed 
which,  on  the  addition  of  2  c.c.  of  hydrochloric  acid 
turns  to  a  violet  blue,  the  sample  contains  formalde- 
hyde. 

Hehner's  test  for  formaldehyde  in  milk,  as  modified  by 
Richmond  and  Bosely,  is  as  follows  :  The  milk  is  diluted 
with  an  equal  bulk  of  water,  and  94  per  cent,  sulphuric 
acid  containing  a  trace  of  a  ferric  salt  is  added  in  such  a 
way  that  it  will  form  a  layer  under  the  milk  ;  in  the 
presence  of  formaldehyde,  a  violet  coloration  is  formed  at 
the  junction  of  the  layers  ;  in  the  absence  of  formalde- 

19—2 


292      Industrial  Organic  Analysis 

hyde,  only  a  faint,  greenish  coloration  will  be  observed. 
This  test,  which  is  of  extreme  delicacy,  is  only  applicable 
to  milk.  Further  tests  for  formaldehyde,  applicable  to 
other  foodstuffs,  are  given  below. 

In  addition  to  the  objections  to  the  use  of  formalde- 
hyde in  milk,  akeady  mentioned,  it  may  be  pointed  out 
that  this  preservative  has  been  shown  by  Sommerfeld  to 
have  a  more  pronounced  action  on  the  relatively  harmless 
organisms,  e.g.,  the  lactic  acid  producing  bacteria,  than 
on  the  pathogenic  and  putrefactive  organisms  which  may 
occur  in  milk.  This  was  found  to  be  the  case  when 
formaldehyde  was  added  in  the  proportion  of  i  to  10,000. 
It  is  obvious  that  if  the  activity  of  the  harmless  bacteria 
is  impaired  to  a  greater  degree  than  that  of  the  more 
injurious  organisms,  the  latter  will  be  able  to  develop 
more  freely  and  render  the  milk  unfit  for  use.  (Compare 
Chapter  VI. ,  p.  215.) 

The  Detection  of  Hydrogen  Peroxide  in  Milk. — Like 
formaldehyde,  this  preservative  is  gradually  decomposed 
by  milk,  and  disappears  after  a  time.  The  hydrogen 
peroxide  is  decomposed  into  water  and  oxygen  by  a 
catalytic  action,  the  nascent  oxygen  having  a  bactericidal 
action.  This  principle  is  utilised  in  Budde's  process  for 
sterilising  milk.  Hydrogen  peroxide  is  usually  added  in 
the  proportion  of  about  0-05  per  cent.  ;  sometimes  the 
milk  is  heated  with  the  hydrogen  peroxide  in  closed 
vessels  to  50°  for  8  to  10  hours,  after  which  it  is  sup- 
posed to  be  sterile.  It  has,  however,  been  shown  that 
the  hydrogen  peroxide  does  not  destroy  all  the  patho- 
genic organisms  when  used  in  the  concentration  men- 
tioned, while  if  the  amount  necessary  for  complete 
sterilisation,  i.e.,  0-4  per  cent.,  is  added,  a  bitter  taste  is 
produced,  probably  owing  to  the  oxidation  of  the  olein 
of  the  milk  fat. 

Hydrogen  peroxide  may  be  detected  by  adding  to 
about  10  c.c.  of  the  milk  a  few  drops  of  a  recently 
prepared  solution  of  potassium  iodide  and  starch,  and 
then  a  very  small  quantity  of  dilute  ferrous  sulphate 


Preservatives,  etc.,  in  Foods     293 

solution  ;  in  the  presence  of  hydrogen  peroxide,  a  blue 
coloration  will  be  developed.  The  reaction  is  extremely 
delicate. 

The  Detection  of  Sodium  Fluoride. — The  milk,  cream 
or  aqueous  serum  from  butter  or  margarine  is  made 
distinctly  alkaline  with  milk  of  lime,  evaporated  to 
dryness  and  incinerated.  The  ash  is  placed  in  a  platinum 
dish,  moistened  with  water,  and  5  c.c.  of  concentrated 
sulphuric  acid  are  added.  The  dish  is  then  covered  with 
a  watch  glass  which  has  been  coated  on  the  under  side 
with  paraffin  wax,  the  latter  having  been  scraped  off  in 
a  few  places.  Gentle  heat  is  applied,  care  being  taken 
not  to  melt  the  paraffin.  If  the  exposed  parts  of  the 
glass  become  etched  in  the  course  of  about  half  an  hour, 
the  presence  of  fluoride  may  be  inferred.  In  the  presence 
of  compounds  of  boron,  the  test  must  be  modified  as 
described  under  the  next  heading. 

The  Detection  of.  Fluoride  in  presence  of  Boron  Com- 
pounds in  Butter  and  Margarine. — The  test  for  fluoride 
is  complicated  by  the  presence  of  boric  acid  or  borax, 
owing  to  the  formation  of  fluoboric  acid,  which  has  no 
etching  effect  on  glass.  As  many  butters  and  margarines 
contain  boric  acid  or  borax,  Otto  and  C.  W.  Hehner  have 
devised  the  following  test  :  50  grams  of  the  sample  are 
melted  and  mixed  with  50  c.c.  of  hot  water;  the  aqueous 
layer  is  separated  from  the  fat  in  a  separating  funnel 
and  made  alkaline  with  sodium  carbonate.  After  adding 
an  excess  of  calcium  chloride,  the  liquid  is  evaporated 
to  dryness  ;  the  residue  is  incinerated,  and  the  ash 
extracted  with  dilute  acetic  acid,  transferring  it  to  a 
filter  and  washing  with  the  acid.  The  boron  com- 
pounds are  thus  dissolved,  while  the  calcium  fluoride 
remains  on  the  filter.  The  filter  with  the  insoluble 
ash  is  transferred  to  a  platinum  dish,  dried  and 


294      Industrial  Organic  Analysis 

incinerated.     The    ash   is   then   tested   for  fluoride   as 
described  above. 

PRESERVATIVES  IN  MEAT,  WINE  AND  OTHER  FOODS. 

Most  of  the  methods  for  the  identification  and  estima- 
tion of  preservatives  given  above  may  be  applied  in  the 
case  of  other  foods,  the  method  of  extracting  the  pre- 
servative being  varied  to  suit  any  particular  case.  Solid 
or  semi-solid  foods,  such  as  meat  or  sausages,  are  usually 
comminuted  and  extracted  with  water  ;  the  filtered 
aqueous  extract  may  then  be  acidified  and  extracted 
with  ether  or  chloroform,  when  such  preservatives  as 
benzoic  and  salicylic  acids,  or  saccharin  will  pass  into 
the  organic  solvent,  while  fluorides,  boric  acid,  etc.,  will 
remain  in  solution  in  the  aqueous  layer.  In  some  cases 
it  may  be  convenient  to  render  the  aqueous  extract 
alkaline  and  to  evaporate  it  to  a  small  bulk  before 
acidifying  and  extracting  with  ether.  Liquids  may  be 
treated  in  the  same  way  as  the  aqueous  extract  from  solid 
foods.  Another  general  method  is  to  steam  distil  the 
material  in  presence  of  phosphoric  acid,  when  benzoic 
and  salicylic  acids,  etc.,  and  sulphurous  acid,  from 
sulphites,  will  pass  over  with  the  distillate. 

As  examples  of  the  methods  generally  in  use,  the 
detection,  and  in  some  cases  also  the  estimation,  of  some 
of  the  commoner  preservatives  in  meat  and  wine  will  be 
described.  Besides  common  salt,  nitre,  sugar,  wood 
smoke,  vinegar  and  spices,  the  commonest  preservatives 
used  in  meat  and  meat  food  products  are  formaldehyde, 
boric  acid  or  borax,  and  sulphites  (usually  sodium  or 
calcium  bisulphites).  The  use  of  the  latter  preserva- 
tives in  meat  is  forbidden  by  law  in  Germany  and  the 
United  States.  American  packers  may,  however,  under 
the  direction  of  the  foreign  purchaser  or  his  agent,  add 
preservatives  to  meat  and  meat -food  products  intended 
for  export,  in  proportions  which  do  not  conflict  with  the 


Preservatives,  etc.,  in   Foods     295 

laws  of  the  foreign  country  to  which  they  are  to  be 
exported.  As  has  already  been  pointed  out,  the  law  of 
the  United  Kingdom  contains  no  definite  prohibitions 
for  the  use  of  preservatives  in  foods,  except  in  the  case  of 
milk  and  cream.  It  may  be  mentioned  that  a  conviction 
for  29  grains  of  boric  acid  per  pound  in  sausages  was 
quashed  under  section  6  of  the  Food  and  Drugs  Act  of 
1875,  while  prosecutions  have  succeeded  for  40  grains  of 
boric  acid  per  pound,  and  larger  quantities,  in  similar 
foodstuffs.  Formaldehyde  is  generally  held  to  be  harm- 
ful ;  it  is  used  for  fumigating  meat  intended  for  long 
transit,  and  is  a  very  efficient  preservative,  but  has  a 
harmful  effect  on  the  digestion. 

The  Departmental  Committee  on  Preservatives  and 
Colouring  Matters  in  Foods,  1901,  recommended  that  the 
use  of  formaldehyde  or  its  preparations  in  foods  or  drink.3 
be  absolutely  prohibited.  In  a  report  to  the  Local 
Government  Board,  1909,  Dr.  Buchanan  recommended 
that  meat  traders  and  importers  should  consider  the 
practicability  of  limiting  the  use  of  formaldehyde  to  the 
adequate  disinfection  of  the  holds  in  which  the  meat  is  to 
be  conveyed,  before  it  is  introduced.  In  a  report  to  the 
Local  Government  Board  on  Preservatives  in  Meat 
Foods  packed  in  Cans  or  Glasses,  1908,  Dr.  MacFadden 
points  out  that  a  considerable  proportion  of  the  samples 
of  canned  meat  foods,  both  of  British  and  American 
manufacture,  examined  by  different  analysts,  contained 
either  boron  or  sulphite  preservatives.  In  reviewing  he 
recommends  that  "  steps  should  be  taken  to  secure  that 
specified  chemical  preservatives  should  not  be  used  in 
the  preparation  of  canned  meats  intended  for  use  in  this 
country,"  and  that  "  in  any  schedule  of  prohibited 
preservatives,  boron  compounds,  sulphites  and  prepara- 
tions of  sulphurous  acid,  benzoic  acid  and  formalin 
should  be  included." 

The  chief  preservative  used  in  wines,  beers  and  other 
beverages,  is  salicylic  acid.  The  Departmental  Com- 
mittee on  Preservatives  and  Colouring  Matters  in  Foods, 
1901,  recommended  that  this  preservative  should  not  be 
used  in  a  greater  proportion  than  i  grain  per  pint  in 
liquid  food,  and  I  grain  per  pound  in  solid  food.  Quite  a 


296      Industrial  Organic  Analysis 

number  of  prosecutions  have  been  successfully  main- 
tained against  salicylic  acid,  though  the  results  show  but 
little  uniformity  ;  thus  cases  may  be  cited  in  which 
prosecutions  for  13  grains  per  pint  in  ginger  wine  failed, 
and  7-2  grains  per  pint  in  similar  material  procured 
conviction.  A  prosecution  for  17  grains  per  Ib.  in  jam 
failed,  while  2-67  grains  per  Ib.  procured  conviction. 
Besides  salicylic  and  benzoic  acids,  beverages  may  be 
preserved  with  sulphites,  fluorides  and  boric  acid. 
Saccharin  is  sometimes  added  to  wines  and  sweet 
beverages  in  order  to  reduce  the  amount  of  sugar  and 
lessen  the  fermentation  ;  it  is  said  to  conceal  inferior 
quality,  and  also  to  have  a  harmful  effect  on  digestion. 
Fruit  juices,  jams,  etc.,  may  be  preserved  with  benzoic, 
salicylic,  boric  and  formic  acids. 

%  In  Germany,  practically  all  the  preservatives  men- 
tioned above  (and  in  the  case  of  acids,  also  their  salts)  are 
forbidden  by  law  in  meat  and  wine.  Sulphurous  acid 
and  sulphites,  not  being  considered  poisonous,  are  not 
excluded  from  wine.  They  are,  however,  prohibited  in 
meat,  as  sulphurous  acid  tends  to  restore  the  colour  of 
bad  meat.  In  France,  most  of  the  common  preserva- 
tives are  forbidden  in  wine,  while  the  use  of  salicylic  acid, 
benzoic  acid,  and  their  salts,  and  formaldehyde  as 
preservatives  is  forbidden  altogether. 

Detection  and  Estimation  of  Formaldehyde  in  Meat. — 
The  reaction  given  above  for  the  detection  of  formalde- 
hyde in  milk,  depending  on  the  formation  of  a  violet 
coloration  in  the  presence  of  proteins,  mineral  acid  and 
an  oxidising  agent,  cannot  be  applied  here,  as  meat  gives 
a  violet  colour  on  warming  with  mineral  acid  in  the 
absence  of  the  aldehyde.  The  following  method  has 
been  devised  by  Dr.  Schryver  (Report  to  Local  Govern- 
ment Board  on  the  Application  of  Formaldehyde  to  Meat, 
1909)  for  the  detection  of  formaldehyde,  polymerised 
formaldehyde  or  formaldehyde  which  has  entered  into 
combination  with  other  substances,  i.e.,  the  proteins  of 
the  meat :  10  grams  of  the  minced  meat  are  heated  for 


Preservatives,  etc.,  in  Foods     297 

5  minutes  on  a  boiling  water  bath,  with  water  to  every 
10  c.c.  of  which  have  been  added  2  c.c.  of  a  I  per  cent, 
solution  of  phenyl  hydrazine  hydrochloride.  The  quan- 
tity of  liquid  is  varied  according  to  the  amount  of 
formaldehyde  present.  In  most  cases  where  the  amount 
of  formaldehyde  is  i  part  in  50,000  or  less,  10  c.c.  of 
water  and  2  c.c.  of  phenyl  hydrazine  solution  are  em- 
ployed. Where  the  concentrations  are  higher,  larger 
quantities  of  the  liquid  must  be  used.  Thus,  where  the 
concentration  of  the  aldehyde  in  the  meat  reaches  i  part 
in  5,000,  10  grams  of  meat  are  heated  with  100  c.c.  of 
water  and  20  c.c.  of  i  per  cent,  phenyl  hydrazine  hydro- 
chloride  solution.  After  heating,  the  liquid  is  cooled 
and  filtered  from  the  coagulum  through  a  loose  plug  of 
cotton  wool.  To  12  c.c.  of  the  filtrate  are  added  i  c.c.  of 
a  5  per  cent,  solution  of  potassium  ferricyanide  and  4  c.c. 
of  concentrated  hydrochloric  acid  for  each  12  c.c.  of 
water  and  phenyl  hydrazine  reagent  employed  in  the 
test.  In  the  presence  of  formaldehyde,  a  brilliant 
fuschine-like  colour  is  developed,  which  reaches  its  full 
intensity  after  a  few  minutes'  standing,  and  keeps  without 
marked  deterioration  for  several  hours.  By  comparison 
of  the  colour  with  standard  solutions  containing  known 
amounts  of  formaldehyde,  the  amount  of  formaldehyde 
in  the  meat  sample  can  be  ascertained. 

Detection  and  Estimation  of  Boric  Acid  (or  Borax), 
(a)  In  Meat. — The  following  is  the  German  official 
method  for  meat  inspection :  5t)  grams  of  the  comminuted 
meat  are  triturated  with  50  c.c.  of  water,  to  which  has 
been  added  0-2  per  cent,  of  concentrated  hydrochloric 
acid  (specific  gravity  1-124).  The  mixture  is  allowed  to 
stand  in  a  beaker  for  half  an  hour,  after  which  it  is  heated 
on  a  boiling  water  bath  for  half  an  hour,  with  occasional 
stirring,  the  beaker  being  covered  with  a  watch  glass. 


298      Industrial  Organic  Analysis 

The  warm  mass  is  pressed  in  muslin,  and  the  liquid 
extract  poured  through  a  moist  filter.  The  filtrate  is 
made  faintly  alkaline  to  phenol  phthalein  by  the  addition 
of  decinormal  sodium  hydroxide  solution,  and  evaporated 
to  25  c.c.  5  c.c.  of  the  liquid  thus  obtained  are  acidified, 
filtered  and  tested  with  turmeric  paper  as  follows  :— 

A  strip  of  turmeric  paper  8  cm.  long  and  I  cm.  wide  is 
wetted  half  its  length  with  the  acid  liquid  and  dried  on 
a  watch  glass  at  60°  to  70°.  If  no  change  is  observed  on 
the  part  which  was  wetted,  then  boric  acid  is  absent. 
If  a  red  or  orange-red  colour  is  produced,  a  little  2  per 
cent,  solution  of  sodium  carbonate  (anhydrous)  should 
be  added  ;  if  a  reddish  brown  spot  is  produced,  which 
does  not  differ  from  that  got  with  pure  turmeric  paper 
and  sodium  carbonate,  then  boric  acid  is  absent.  If, 
however,  the  sodium  carbonate  solution  produces  a  blue 
spot;  then  boric  acid  is  present.  If  a  bluish  violet 
coloration  is  produced,  or  the  indications  obtained  are  in 
any  way  doubtful,  the  flame  test  should  be  applied,  the 
official  directions  for  which  are  as  follows  : — 

5  c.c.  of  the  concentrated  alkaline  solution,  obtained 
as  described  above,  are  evaporated  to  dryness  and 
incinerated  in  a  platinum  dish.  The  latter  process  may 
be  greatly  facilitated  by  extracting  the  charred  mass 
with  20  c.c.  of  water,  then  burning  off  the  residual  coal, 
and  finally  returning  the  aqueous  extract  to  the  dish, 
evaporating  to  dryness  and  heating  the  residue  to  120°. 
The  ash  is  well  mixed  with  5  c.c.  of  methyl  alcohol  and 
0-5  c.c.  of  concentrated  sulphuric  acid,  and  the  whole  is 
transferred  to  a  100  c.c.  Erlenmeyer  flask  with  the  aid  of 
a  further  quantity  of  5  c.c.  of  methyl  alcohol.  The  flask 
is  closed  with  a  cork,  and  shaken  at  frequent  intervals 
during  half  an  hour,  after  which  all  the  alcohol  is  dis- 
tilled off,  heating  the  flask  in  a  water  bath  at  80°  to  85°. 


Preservatives,  etc.,  in  Foods     299 

The  distillate  is  introduced  into  a  glass  cylinder  or  test 
tube  of  about  40  c.c.  capacity,  about  6  cm.  high,  and 
fitted  with  a  cork  carrying  two  glass  tubes  bent  at  right 
angles,  one  of  which  passes  to  the  bottom  of  the  vessel, 
and  the  other  just  through  the  cork.  A  current  of 
hydrogen  is  passed  through  the  liquid,  and  ignited  as  it 
emerges  from  the  shorter  tube  ;  if  a  green- edged  flame  is 
produced  then  boric  acid  is  present. 

Estimation  of  Boric  Acid. — This  may  be  carried  out  by 
evaporating  a  definite  proportion  of  the  concentrated 
alkaline  liquid  used  for  the  above  tests  to  dryness, 
incinerating  as  described  under  the  green-flame  test, 
dissolving  the  ash  in  water  and  titrating  the  liquid 
previously  neutralised  to  methyl  orange  with  deci- 
normal  sodium  hydroxide  solution  in  presence  of  glycerol 
and  phenol  phthalein,  as  described  on  p.  288. 

(b)  In  Wine,  Fruit  Juices,  Jams,  etc. — The  following 
method  is  due  to  Allen  and  Tankard  :  100  c.c.  of  the 
liquid  are  evaporated  to  dryness  with  10  c.c.  of  a  10  per 
cent,  solution  of  calcium  chloride.  In  the  case  of  solid 
or  semi-solid  material,  such  as  jams,  mincemeat,  etc.,  the 
mass  should  be  broken  up  and  the  calcium  chloride 
solution  well  mixed  with  it.  The  residue  is  incinerated  by 
first  charring,  then  extracting  the  mass  with  150  c.c.  of 
water,  and  filtering  the  aqueous  extract  from  the  coal, 
which  is  burnt  off  by  itself.  The  residue  thus  obtained 
is  boiled  with  a  second  portion  of  150  c.c.  of  water, 
allowed  to  stand  for  12  hours  and  filtered  cold.  The 
filtered  liquid  is  united  with  that  previously  obtained 
from  the  charred  mass,  and  the  boric  acid  in  solution 
estimated  by  titration  as  described  on  p.  288.  The 
incinerated  residue  may  be  extracted  with  a  further 
150  c.c.  of  water,  and  the  filtrate  thus  obtained  titrated 
for  boric  acid,  to  make  sure  that  this  constituent  has  been 


300      Industrial  Organic  Analysis 

completely  extracted.  The  qualitative  tests  for  boric 
acid  may  be  carried  out  as  described  under  the  preceding 
heading  (Detection  of  Boric  Acid  in  Meat)  on  a  portion 
of  the  aqueous  extract  previously  made  alkaline,  and 
concentrated  or  evaporated  to  dryness  as  the  case  may 
require. 

Detection  and  Estimation  of  Sulphurous  Acid  or 
Sulphites  in  Meat,  Wine,  etc. — The  following  are  the 
German  official  methods  for  the  detection  and  estimation 
of  sulphurous  acid  and  sulphites  in  meat  and  wine  ;  they 
may  also  be  applied  to  other  materials,  such  as  jams, 
cider,  beer,  etc.  : — 

The  qualitative  test  is  carried  out  as  follows  with  meat  : 
30  grams  of  the  comminuted  meat  are  rapidly  mixed 
with  5  c.c.  of  a  25  per  cent,  solution  of  phosphoric  acid 
in  a  100  c.c.  Erlenmeyer  flask.  The  latter  is  closed  with 
a  cork  in  which  a  slit  has  been  cut,  so  that  a  piece  of 
potassium  iodate  and  starch  paper  may  be  suspended 
from  it.  The  test  paper  should  be  freshly  prepared  by 
soaking  filter  paper  in  a  mixture  in  equal  parts  of  i  per 
cent,  solutions  of  potassium  iodate  and  starch,  drying  at 
a  gentle  heat  and  cutting  into  strips  of  convenient  size. 
A  strip  of  this  paper  is  suspended  so  that  the  end  is 
about  i  cm.  above  the  meat,  and  I  cm.  of  the  lower 
portion  is  moistened  with  water.  If  in  the  course  of 
10  minutes  no  blue  colour  appears  (usually  seen  at  the 
junction  of  the  wet-  and  dry  parts  of  the  paper),  the  cork 
is  loosened  and  the  flask  is  placed  on  the  water  bath, 
warmed,  then  closed  with  the  cork  and  allowed  to  cool ; 
if  no  colour  is  observed  on  the  test  paper  after  half  an 
hour,  it  may  be  assumed  that  the  meat  is  free  from  sul- 
phurous acid  or  sulphites. 

The  method  recommended  for  the  estimation  is  as 
follows  :  30  grams  of  the  comminuted  meat,  or  100  c.c.  of 


Preservatives,  etc.,  in  Foods     301 

wine,  are  mixed  with  sufficient  sodium  carbonate  solution 
to  render  the  whole  alkaline,  in  a  500  c.c.  round-bottomed, 
long-necked  flask,  the  volume  of  liquid  being  made  up 
to  about  150  c.c.  After  standing  for  I  hour,  the  flask 
is  connected  up  as  for  a  steam  distillation,  on  the  one 
side  with  an  inlet  tube  passing  well  below  the  surface  of 
the  contents,  and  on  the  other  side  with  a  Liebig  con- 
denser which  is  connected  at  the  other  end  by  means  of 
an  adapter,  with  a  U  tube  having  3  bulbs  (Peligot  tube), 
which  must  be  capable  of  holding  150  c.c.  of  liquid  while 
gas  is  being  passed  through  it.  A  stream  of  carbon 
dioxide,  entering  through  the  inlet  tube  in  the  flask,  is 
passed  through  the  whole  apparatus  ;  when  all  the  air 
has  been  displaced,  the  Peligot  tube  is  charged  with 
50  c.c.  of  a  solution  prepared  by  dissolving  7-5  grams  of 
potassium  iodide  and  5  grams  of  iodine  in  i  litre  of  water 
(pure  materials  being  used,  so  that  the  solution  is  free 
from  sulphates),  and  10  c.c.  of  a  25  per  cent,  solution  of 
phosphoric  acid  are  added  to  the  contents  of  the  flask, 
the  cork  being  removed  and  replaced  as  quickly  as  pos- 
sible ;  the  current  of  carbon  dioxide  is  maintained 
throughout.  After  connecting  up  the  apparatus  as  before, 
about  half  of  the  liquid  is  distilled  off,  still  maintaining 
the  current  of  carbon  dioxide.  All  the  sulphurous  acid 
will  now  have  been  driven  over  into  the  Peligot  tube, 
where  it  will  be  oxidised  to  sulphuric  acid  by  the  iodine 
solution,  which  should  remain  brown  throughout,  showing 
an  excess  of  iodine  to  be  present.  The  contents  of  the 
Peligot  tube  and  rinsings  are  transferred  to  a  beaker, 
hydrochloric  acid  and  barium  chloride  are  added,  the 
mixture  boiled,  the  barium  sulphate  being  precipitated, 
collected,  washed  and  weighed  as  in  an  ordinary  sulphate 
determination.  The  barium  sulphate  may  then  be 
calculated  to  sulphur  dioxide  or  sulphurous  acid.  As 


302      Industrial  Organic  Analysis 

already  mentioned,  the  above  methods  may  be  applied 
or  adapted  to  materials  other  than  wine  or  meat.  Some 
analysts  use  a  definite  volume  of  iodine  solution  of  known 
strength,  determined  by  titration  with  standard  sodium 
thiosulphate  solution,  and  estimate  the  iodine  which  has 
been  used  up  in  the  oxidation  of  the  sulphurous  acid 
by  a  second  titration  with  thiosulphate  solution.  The 
latter  method  is  somewhat  more  expeditious  than  the  one 
given  above. 

Detection  of  Fluorides,  (a)  In  Meat. — The  German 
official  method  is  as  follows  :  25  grams  of  the  minced 
meat  are  thoroughly  mixed  with  an  excess  of  milk  of 
lime  in  a  platinum  dish,  dried  and  incinerated.  The 
residue  is  treated  with  3  drops  of  water  and  i  drop  of 
concentrated  sulphuric  acid,  the  etching  test  for  hydro- 
fluoric acid  being  applied  in  the  usual  way,  as  described 
on  p.  293. 

(b)  In    Wine. — -Vandam    recommends    the    following 
method  :   To   100   c.c.   of  the  sample  in   a  measuring 
cylinder  are  added  0-5  to  i  c.c.  of  a  20  per  cent,  sodium 
sulphate  solution  and  10  c.c.  of  a  10  per  cent,  barium 
acetate  solution  ;    after  shaking  well,   the  mixture  is 
allowed  to  stand  overnight  and  the  clear  liquid  syphoned 
off.     The  sediment  is  shaken  up  with  100  c.c.  of  hot  water 
and  allowed  to  settle  ;   after  removing  the  clear  liquor, 
the  process  is  repeated  with  a  further  quantity  of  50  c.c. 
of  hot  water.     The  washed  sediment,   containing  the 
insoluble  barium  fluoride,  is  transferred  to  a  double  filter 
and,  when  dry,  incinerated.     The  ash  is  moistened  with 
water,  treated  with  5  c.c.  of  concentrated  sulphuric  acid, 
and  the  etching  test  applied  as  described  on  p.  293.     It 
should  be  noted  that  many  wines  contain  small  traces 
of  fluorides  as  a  natural  constituent. 

(c)  In  Beer. — The  following  method  is  given  in  Allen's 


Preservatives,  etc.,  in  Foods     303 

"  Commercial  Organic  Analysis,"  Vol.  L,  1909  ed.  :  looc.c. 
of  the  sample  are  made  slightly  alkaline  with  ammonium 
carbonate,  boiled,  treated  with  2  to  3  c.c.  of  a  10  per  cent, 
calcium  chloride  solution  and  boiled  again  for  5  minutes. 
The  precipitate  is  filtered  off,  washed,  dried  and  tested 
for  fluoride  in  the  usual  way. 

Detection  and  Estimation  of  Chlorides  and  Nitrates  in 
Meat. — The  following  method,  partly  due  to  Given,  is 
described  by  Leffmann  and  Beam  in  their  "  Food 
Analysis  "  :  It  will  be  necessary  first  to  determine  the 
chlorides  present,  as  these  interfere  with  the  nitrate 
determination  ;  this  may  be  done  by  titrating  the  solu- 
tion obtained  by  extracting  i  gram  of  the  minced  meat 
with  200  c.c.  of  water,  with  decinormal  silver  nitrate 
solution,  using  potassium  chromate  as  indicator  ;  the 
method  of  titration  is  described  in  most  works  on  quanti- 
tative inorganic  analysis. 

For  the  determination  of  nitrates,  I  gram  of  the 
sample  is  placed  in  a  100  c.c.  flask,  50  c.c.  of  water  are 
added,  and  the  mixture  warmed  in  hot  water  for  20 
minutes,  with  occasional  shaking.  For  each  i  per  cent, 
of  sodium  chloride  found  to  be  present,  3  c.c.  of  a  satu- 
rated solution  of  silver  sulphate  are  added,  then  10  c.c. 
of  basic  lead  acetate,  and  5  c.c.  of  alumina  cream  (see 
p.  226),  shaking  after  each  addition.  The  liquid  is 
made  up  to  100  c.c.,  shaken,  filtered  through  a  dry  fluted 
filter,  the  filtrate  being  returned  till  clear.  20  c.c.  of  the 
filtrate  are  evaporated  to  dryness  on  the  water  bath  in  a 
porcelain  dish,  and  the  residue  is  mixed  with  i  c.c.  of 
phenol  disulphonic  acid,  the  preparation  of  which  is 
described  below  ;  without  applying  heat,  the  acid  is 
stirred  over  the  whole  residue  ;  the  mixture  is  completely 
transferred  to  a  Nessler  tube  by  rinsing  with  water,  and 
the  solution  thus  obtained  made  alkaline  with  ammonia 


304      Industrial  Organic  Analysis 

or  soda.  By  the  interaction  of  the  nitrate,  the  phenol 
disulphonic  acid  and  the  alkali,  an  alkaline  picrate  is 
formed,  the  depth  in  colour  due  to  the  latter  being  pro- 
portional to  the  amount  of  nitrate  present  in  the  sample. 
The  determination  may  therefore  be  made  by  comparison 
with  a  solution  which  has  been  similarly  prepared  by 
evaporating  to  dryness  on  the  water  bath  a  known 
volume,  say,  I  c.c.,  of  a  standard  solution  of  potassium 
nitrate,  containing  o-ooi  gram  of  the  salt  per  cubic  centi- 
metre, treating  with  phenol  disulphonic  acid,  transferring 
to  a  Nessler  tube  by  means  of  water  and  rendering  alka- 
line as  before.  The  two  picrate  solutions  are  made  up 
to  the  same  volume  and  compared.  If  the  difference 
in  the  depth  of  colour  is  not  great,  some  of  the  deeper 
coloured  solution  may  be  poured  off  till  the  tints  observed 
in  the  two  tubes  when  placed  side  by  side  on  a  white 
surface  are  sensibly  equal.  The  relative  depths  of  the 
two  layers  of  liquid  may  then  be  taken  as  a  basis  for 
calculation.  If,  on  the  other  hand,  the  difference  in 
tints  is  very  marked,  another  solution  for  comparison 
must  be  made  from  a  greater  or  smaller  quantity  of 
potassium  nitrate. 

The  phenol  disulphonic  acid  is  prepared  as  follows  : 
37  grams  of  pure  sulphuric  acid  and  3  grams  of  pure 
phenol  are  heated  for  6  hours  in  a  flask  immersed  in 
boiling  water.  The  resulting  reagent  may  crystallise  on 
cooling,  but  can  easily  be  liquefied  on  gentle  warming. 

Detection  and  Estimation  of  Salicylic  and  Benzoic  Acids. 
—The  following  method  is  recommended  by  Harry  and 
Mummery  for  the  detection  and  estimation  of  salicylic 
acid  in  wine  and  beer  ;  it  has  the  advantage  of  elimina- 
ting tannins  which  may  mask  the  reaction  of  salicylic 
acid  with  ferric  salts,  and  also  substances  which  tend  to 
give  rise  to  emulsions  on  extracting  with  immiscible 


Preservatives,  etc.,  in  Foods     305 

solvents.  It  may  equally  well  be  applied  to  solid  or 
semi-solid  foods,  in  which  case  it  will  be  necessary  to 
make  either  an  aqueous  solution  or  a  slightly  alkaline 
aqueous  extract.  As  far  as  the  actual  extraction  of  the 
preservative  from  the  sample  is  concerned,  the  method 
is  also  applicable  to  benzoic  acid. 

100  c.c.  of  the  sample  (or  aqueous  solution  or  extract) 
are  made  alkaline  with  5  c.c.  of  normal  sodium  hydroxide 
solution,  and  the  alcohol  (if  any)  is  driven  off  at  a  tem- 
perature just  below  the  boiling  point.  The  following 
operations,  up  to  the  ether  extraction,  have  for  their 
object  the  removal  of  tannins  and  pectinous  or  albu- 
minous matter  from  the  aqueous  solution  of  the  preserva- 
tive. 5  c.c.  of  normal  hydrochloric  acid  are  added,  and 
then  20  c.c.  of  basic  lead  acetate  solution  ;  the  mixture 
is  made  alkaline  with  about  20  c.c.  of  normal  sodium 
hydroxide  solution,  and  made  up  to  200  c.c.  with  water. 
The  tannins  are  precipitated  while  the  lead  salicylate 
(or  benzoate)  is  soluble  in  the  alkaline  solution.  After 
mixing,  heating  in  boiling  water  and  cooling,  the  liquid 
is  filtered  through  a  dry  filter  ;  100  c.c.  of  the  filtrate  are 
acidified  with  hydrochloric  acid,  which  will  precipitate 
albuminous  matter  together  with  lead  chloride,  besides 
liberating  the.  salicylic  (or  benzoic)  acid.  The  filtrate 
and  washings  from  the  last  precipitation  are  extracted 
three  times  with  ether,  the  ethereal  extracts  united 
and  evaporated  to  dryness.  The  identification  of 
salicylic  or  benzoic  acids  may  be  carried  out  by  the 
methods  already  described  (see  p.  289  and  below).  For 
the  estimation  of  salicylic  acid,  Harry  and  Mummery 
recommend  the  following  method  :  The  acid  is  dissolved 
in  a  small  quantity  of  dilute  alcohol  and  made  up  to 
100  c.c.  in  a  Nessler  tube.  The  colour  produced  with 
ferric  chloride  solution  is  then  compared  with  the  colour 

I.O.A.  20 


306      Industrial  Organic  Analysis 

produced  on  adding  an  equal  amount  of  ferric  chloride 
to  solutions  of  the  same  volume  in  Nessler  tubes  con- 
taining known  amounts  of  salicylic  acid. 

In  the  presence  of  both  benzoic  and  salicylic  acids, 
van  der  Laan  and  Tydens  recommend  the  following 
method  for  their  separation  :  After  estimating  the 
salicylic  acid  colorimetrically  (see  above)  in  a  portion  of 
the  aqueous  solution,  the  rest  of  the  solution  is  extracted 
with  ether,  the  acids  obtained  by  evaporation  of  the 
ethereal  extract  are  dissolved  in  10  to  20  c.c.  of  quarter- 
normal  caustic  alkali  solution,  and  a  slight  excess  of  a 
5  per  cent,  potassium  permanganate  solution  is  added. 
After  gentle  heating  by  means  of  a  small  flame,  the 
excess  of  permanganate  and  the  separated  manganic 
hydroxide  are  reduced  by  adding  a  saturated  solution  of 
sulphur  dioxide.  The  salicylic  acid  will  be  destroyed 
by  -the  oxidising  agent,  while  the  benzoic  acid  will 
remain  unaffected.  The  acidified  solution  may  then 
be  extracted  with  ether,  and  the  benzoic  acid  thus 
obtained  identified,  and  estimated  by  titration.  For 
the  method  of  separating  salicylic  and  benzoic  acids 
from  saccharine,  which  may  mask  the  reactions  for  the 
identification  of  the  former,  see  below. 

Detection  and  Estimation  of  Saccharine  in  Beverages.— 
Saccharine,  or  ortho-benzoyl-sulphone-imide,  or  its 
sodium  salt,  is  not  a  preservative  in  the  true  sense  of  the 
word  ;  it  is  added  to  beverages  in  place  of  sugar,  so  as  to 
reduce  fermentation. 

The  imide  is  removed  on  extracting  the  acidified 
material  with  ether,  and  may  be  detected  by  the  sweet 
taste  of  the  residue  obtained  on  evaporation  of  the  ether. 
If  present  together  with  benzoic  and  salicylic  acids,  it 
will  generally  accompany  these  if  they  are  extracted 
from  the  material  by  means  of  organic  solvents.  Separa- 


Preservatives,  etc.,   in   Foods     307 

tion  may  be  accomplished  by  acidifying  200  grams  of  the 
sample  with  5  c.c.  of  a  20  per  cent,  phosphoric  acid 
solution,  and  distilling  almost  to  dryness  ;  the  acids  will 
be  found  in  the  distillate,  from  which  they  may  be 
obtained  by  acidifying  and  extracting  with  ether,  while 
the  saccharine  will  remain  in  the  flask,  and  may  be 
obtained  by  extracting  the  contents  with  water  and 
extracting  the  acid  solution  thus  obtained  with  ether  in 
the  usual  way. 

Allen's  process  for  the  estimation  of  saccharine  in  beer 
is  as  follows  :  The  beer  is  concentrated  to  one-third  of 
its  original  bulk,  and  if  not  acid,  it  is  rendered  so  by  the 
addition  of  a  little  pure  phosphoric  acid.  The  liquid  is 
extracted  with  ether,  the  latter  evaporated,  and  the 
residue  mixed  with  an  excess  of  anhydrous  sodium 
carbonate  and  a  little  potassium  nitrate,  and  ignited  till 
all  organic  matter  has  been  burnt  off.  A  determination 
of  sulphate  in  the  residue  is  made  by  dissolving  in  water, 
acidifying  with  hydrochloric  acid,  adding  barium 
chloride  solution,  etc.,  as  usual.  The  factor  for  calcu- 
lating the  barium  sulphate  to  saccharine  is  0-785. 
Care  should  be  taken  that  the  reagents  employed  are 
free  from  sulphur  compounds. 

Detection  and  Estimation  of  Formic  Acid. — This  pre- 
servative is  chiefly  used  for  fruit  juices  and  preserves. 
In  honey  it  is  present  as  a  natural  constituent  in  quanti- 
ties up  to  o-2i  per  cent. 

According  to  Croner  and  Seligmann,  formic  acid  is 
separated  from  the  sample  by  mixing  100  grams  of  the 
latter  with  400  c.c.  of  water,  acidifying  with  phosphoric 
acid  and  distilling  in  steam.  500  c.c.  of  distillate  are 
collected,  sufficient  caustic  soda  solution  being  placed  in 
the  receiver  to  keep  the  whole  alkaline.  The  alkaline 
distillate  is  evaporated  to  10  c.c.,  treated  with  an  excess 

20 2 


308      Industrial  Organic  Analysis 

of  baryta  solution  and  filtered  ;  the  excess  of  baryta  is 
then  precipitated  as  sulphate  by  the  addition  of  sulphuric 
acid,  and  the  liquid  again  filtered.  By  this  treatment, 
acids  other  than  formic  acid,  which  might  mask  the  re- 
actions for  the  latter,  are  removed.  The  filtrate  is  boiled 
with  mercuric  chloride  solution,  when  in  the  presence  of 
formic  acid,  a  precipitate  of  mercurous  chloride  will  be 
produced.  According  to  Smith,  formic  acid  may  be 
identified  by  adding  to  the  acid  steam  distillate  a  slight 
excess  of  ammonia  above  that  required  for  neutralisation, 
evaporating  to  a  small  bulk  and  adding  a  few  drops  of 
neutral  ferric  chloride  solution  ;  in  the  presence  of  formic 
or  acetic  acids,  a  red  coloration  will  be  produced  ;  on 
shaking  with  96  per  cent,  alcohol,  a  precipitate  will  be 
produced  in  the  presence  of  formic  acid,  but  not  fn  the 
presence  of  acetic  acid  only.  An  excess  of  acetic  acid 
interferes  with  the  reaction  ;  in  such  a  case,  the  acid 
steam  distillate  is  partially  neutralised  with  about  5  c.c. 
of  normal  soda  solution,  and  concentrated  to  15  c.c.  ; 
most  of  the  formic  acid  will  remain  combined  with  the 
alkali,  while  most  of  the  acetic  acid  will  be  evaporated  off. 
H.  Fincke  recommends  the  following  method  for  the 
estimation  of  formic  acid:  To  the  neutral  or  faintly 
acid  steam  distillate  containing  all  the  formic  acid, 
obtained  as  described  above  (see  also  below),  are  added 
3  to  5  grams  of  sodium  acetate  and  at  least  15  times 
as  much  mercuric  chloride  (in  solution),  by  weight,  as 
there  is  formic  acid  present.  The  mixture  is  heated  for 
2  hours  in  a  flask  fitted  by  means  of  a  rubber  stopper 
with  a  tube  condenser,  immersed  in  boiling  water  so  that 
the  latter  reaches  to  the  level  of  the  liquid  in  the  flask. 
The  mercuric  chloride  solution  is  prepared  by  dissolving 
100  grams  of  mercuric  chloride  and  30  grams  of  sodium 
chloride  in  100  c.c.  of  water.  The  precipitated  mer- 


Preservatives,  etc.,  in  Foods     309 

curous  chloride  is  filtered  off  on  a  Gooch  crucible,  washed 
with  hot  water,  alcohol  and  ether,  dried  and  weighed. 
The  factor  for  converting  mercurous  chloride  into  formic 
acid  is  0-0977. 

As  regards  the  steam  distillation,  it  will  be  necessary 
to  distil  at  least  50oc.c.,in  order  to  bring  practically  all 
the  formic  acid  over.  As  the  estimation  of  the  formic 
acid  depends  on  the  reducing  action  of  the  steam  dis- 
tillate, errors  may  be  introduced  through  the  presence 
of  volatile  aldehydes,  which  may  exist  as  such  in  the 
sample  or  be  produced  from  tartaric  and  other  acids. 
To  obviate  this,  the  steam  may  be  led  through  two  flasks 
containing  a  suspension  of  calcium  carbonate,  before  it 
enters  the  condenser  ;  if  the  flasks  are  kept  heated,  the 
aldehydes  will  pass  on  with  the  steam,  while  the  formic 
acid  is  retained  in  the  flasks  as  calcium  formate.  After 
filtering  off  the  calcium  carbonate  and  washing  with 
water,  the  formic  acid  is  estimated  in  the  filtrate  as 
described  above.  Care  should  be  taken  to  avoid  spray 
being  carried  over  in  the  distillation,  as  reducing  sugars 
may  be  present  in  the  sample,  and  these  would  vitiate 
the  result  by  reducing  some  of  the  mercuric  chloride. 
If  sulphurous  acid  is  present,  the  concentrated  neutral 
distillate,  or  the  filtrate  from  the  calcium  carbonate,  is 
treated  with  about  5  c.c.  of  quarter-normal  caustic  soda 
solution  and  5  c.c.  of  concentrated  hydrogen  peroxide 
solution,  and  left  for  4  hours  ;  the  sulphite  will  then  be 
oxidised  to  sulphate.  A  little  freshly  precipitated 
mercuric  oxide,  made  into  a  paste  with  water,  is  then 
added,  and  after  half  an  hour  the  liquid  is  filtered,  and  the 
formic  acid  estimated  in  the  filtrate  as  described  above. 
In  the  presence  of  salicylic  acid,  common  salt  should  be 
dissolved  in  the  distillate  ;  the  salicylic  acid  will  then 
have  no  influence  on  the  estimation. 


3io      Industrial  Organic  Analysis 

Other  Preservatives. — The  detection  and  estimation  of 
most  of  the  commonly  occurring  preservatives  have 
been  described.  One  or  two  methods  for  detecting  a  few 
other  substances  which  may  be  used  as  preservatives 
are  briefly  outlined  below  :— 

For  the  detection  of  ft  naphfhol  and  other  similar 
substances  the  American  Association  of  Official  Agri- 
cultural Chemists  recommend  that  200  grams  of  the 
acidified  sample  be  distilled  in  steam,  and  the  first 
200  c.c.  of  the  distillate  extracted  with  20  c.c.  of  chloro- 
form ;  on  separating  the  latter,  rendering  it  alkaline 
with  caustic  potash  and  heating  almost  to  boiling  for  a 
few  minutes,  colour  changes  will  occur  as  follows  :  in  the 
presence  of  phenol,  light  red  to  brown,  to  yellow,  to 
colourless.  In  the  presence  of  salol,  light  red,  and  in  the 
presence  of  /3  naphthol,  deep  blue  to  green,  to  brown. 

PART  II. — ARTIFICIAL  COLOURING  MATTERS. 
INTRODUCTORY. 

The  colouring  matters  used  in  foods  may  be  classed 
under  three  headings  :  coal-tar  dyes,  naturally  occurring 
organic  colour^,  and  metallic  or  inorganic  colours.  The 
most  frequently  used  are  the  coal-tar  dyes,  the  majority 
of  which  are  generally  held  to  be  harmless,  especially  in 
the  small  amounts  in  which  they  are  used,  provided,  of 
course,  that  they  are  pure,  and  free  from  arsenic.  The 
great  majority  of  the  vegetable  colouring  matters  are 
also  harmless,  but  the  metallic  colouring  matters,  such 
as  chromates,  copper  salts,  etc.,  are  mostly  injurious, 
even  in  small  amounts.  The  vegetable  colouring  matters 
such  as  cochineal,  anatto,  turmeric  and  saffron,  have 
largely  been  superseded  by  the  coal-tar  colours,  while 
mineral  colours,  with  the  exception,  perhaps,  of  copper 
sulphate,  are  of  comparatively  rare  occurrence  in  foods. 
Among  the  few  poisonous  organic  colours  may  be 
mentioned  gamboge  and  picric  acid. 


Preservatives,  etc.,  in  Foods     311 

The  law  of  the  United  Kingdom  does  not  definitely 
forbid  the  use  of  colouring  matters  in  any  food,  or  limit 
the  amounts  in  which  they  may  be  used ;  as  in  the  case 
of  preservatives  the  addition  of  any  objectionable 
colouring  matter  would,  in  all  probability,  be  dealt  with 
under  section  6  of  the  Sale  of  Foods  and  Drugs  Act  of 
1875  (see  p.  282).  Even  when  the  colouring  matter 
itself  is  harmless,  its  presence  would  be  objectionable  if 
it  had  been  added  in  order  to  conceal  the  inferior  quality 
of  the  food,  as  might  be  done,  for  example,  in  the  case  of 
meat,  milk  or  wine. 

The  Departmental  Committee  on  Preservatives  and 
Colouring  Matters  in  Foods,  1901,  recommended  that  the 
use  of  colouring  matters  of  any  kind  in  milk  offered  for 
sale  in  the  United  Kingdom  should  be  considered  an 
offence  under  the  Sale  of  Foods  and  Drugs  Act.  Further, 
that  the  use  of  copper  salts  in  the  so-called  greening  of 
preserved  foods  should  be  prohibited.  These  recom- 
mendations have,  however,  not  resulted  in  any  legislation 
on  the  subject. 

The  use  of  copper  salts  in  foods  is  prohibited  in  Ger- 
many, and  Austria-Hungary.  In  the  United  States 
certain  specified  colouring  matters  only  are  allowed  in 
foods,  pending  further  enquiry.  Copper  salts  are  not 
prohibited,  but  their  presence  must  be  notified. 

The  identification  of  organic  colouring  matters, 
especially  coal-tar  dyes,  in  foods,  may  often  be  a  difficult 
task,  requiring  considerable  experience.  In  many  cases 
the  full  identification  of  the  colouring  matter  will  not  be 
necessary  ;  it  will  be  sufficient  to  determine  its  nature, 
and  to  be  able  to  say  whether  it  is  harmless  or  injurious. 
By  the  methods  given  below,  the  general  nature  of  the 
colouring  matter  present  may  be  determined ;  the 
methods  for  the  detection  of  some  of  the  commoner 
vegetable  colouring  matters  are  also  given. 

Dyeing  Method  for  the  Detection  of  Coal-tar  Colours. — 
The  method  described  is  recommended  by  Thresh 
and  Porter  ("  Preservatives  in  Food  and  Food  Exami- 
nation ")  for  the  detection  of  both  acid  and  basic  coal- 


312      Industrial  Organic  Analysis 

tar  dyes,  being  based  on  the  method  of  Sostegni  and 
Carpenter!  for  the  detection  of  acid  dyes.  It  may 
generally  be  employed  for  the  detection  of  coal-tar  dyes l 
in  meat,  wine,  confectionery,  milk,  fruit  juices  and 
extracts,  etc. 

The  majority  of  the  dyes  found  in  foods  are  of  an  acid 
nature,  only  few  basic  colours  being  met  with.  If  clean 
white  wool  is  immersed  in  an  acid  solution  of  an  acid 
dye,  or  an  alkaline  solution  of  a  basic  dye,  it  will  take  up 
the  colour  in  both  cases,  forming  insoluble  compounds 
with  dyes  of  both  classes,  presumably  owing  to  its  con- 
taining both  basic  and  acid  constituents.  In  addition  to 
coal-tar  colours,  the  wool  will  take  up  some  vegetable 
colours,  such  as  cochineal  or  logwood  ;  if,  however,  the 
acid  dye,  for  example,  be  dissolved  from  the  wool  by 
means  of  an  alkaline  solution,  then,  on  acidification  of 
the  iatter,  a  .second  piece  of  wool  may  be  dyed  from  it  if 
a  coal-tar  colour  be  present,  but  not  if  the  first  dyeing 
was  due  to  a  vegetable  colour.  The  same  holds  good, 
mutatis  mutandis,  as  regards  basic  dyes. 

In  the  first  place,  it  will  be  necessary  to  prepare  a 
clear  solution  or  extract  of  the  material  containing  the 
colour.  Solid  material  should  be  broken  up  by  passing 
through  a  sausage  grinder,  when  the  colour  may  be 
extracted  by  warming  with  water  or  80  per  cent,  alcohol. 
The  filtered  solution  of  the  colour  is  divided  into  two 
portions  of  50  to  100  c.c.  each,  the  one  of  which  is 
rendered  faintly  alkaline  with  ammonia,  and  the  other 
distinctly  acid  with  hydrochloric  acid.  Into  each  of 
these  solutions  is  put  about  a  foot  of  white  worsted  which 
has  previously  been  boiled  in  a  very  dilute  solution  of 
caustic  soda  in  distilled  water,  and  washed  till  free  from 
alkali.  The  solutions  are  kept  at  the  boiling  point  for 

l  See  also  p.  315. 


Preservatives,  etc.,  in  Foods     313 

an  hour,  or  less  if  the  wool  in  one  of  them  is  distinctly 
dyed.  The  dyed  wool  is  removed,  pressed  between 
sheets  of  niter  paper,  and  washed  by  immersing  in  two 
successive  portions  of  20  c.c.  of  boiling  water. 

The  wool,  if  dyed  from  the  acid  solution,  is  then 
immersed  in  about  20  c.c.  of  a  boiling  solution  of  dilute 
ammonia  containing  10  per  cent,  by  volume  of  the  con- 
centrated ammonia  of  specific  gravity  0-880,  or,  if  dyed 
from  the  alkaline  solution,  in  the  same  quantity  of  boiling 
5  per  cent,  acetic  acid.  The  wools  are  removed,  and  the 
alkaline  liquid  made  acid  by  the  addition  of  acetic  acid, 
and  the  acid  liquid  made  alkaline  by  ammonia.  A  fresh 
piece  of  white,  grease-free  worsted,  about  2  to  3  inches 
long,  is  placed  in  each  solution  ;  after  heating  for  half  an 
hour  on  a  boiling-water  bath,  the  wools  are  removed  and 
washed  in  distilled  water  ;  if  a  coal-tar  dye  was  present 
in  the  original  solution,  one  of  the  samples  should  be 
brilliantly  dyed  ;  if  the  dye  present  is  basic,  the  most 
distinct  dyeing  will  take  place  from  the  alkaline  solutions, 
if  acid,  from  the  acid  solutions.  In  presence  of  vegetable 
colours,  the  second  piece  of  worsted  will  acquire,  at  most, 
a  dirty  appearance. 

For  identification,  the  colour  may  be  removed  from 
the  wool  used  in  a  first  dyeing,  as  described  above,  either 
by  means  of  acid  or  alkali,  and  the  tests  applied  to  the 
solution.  The  identification  of  coal-tar  colours  may  be  a 
difficult  task,  requiring  some  special  experience  ;  the 
subject  is  fully  dealt  with  in  Vol.  V.  of  Allen's  "  Commer- 
cial Organic  Analysis,"  where  several  recognised  schemes 
are  given  ;  the  other  works  mentioned  at  the  end  of  this 
chapter  may  also  be  consulted. 

Colouring  Matters  in  Milk. — According  to  Wynter 
Blyth,  the  addition  of  colours  to  milk  is  frequently 
practised  in  the  United  Kingdom,  the  chief  colours  used 


314      Industrial  Organic  Analysis 

being  sulphonated  azo-dyes  and  ariatto  ;  other  colours 
said  to  have  been  used  are  turmeric,  saffron,  carrot, 
marigold  and  chromates. 

Coal-tar  colours  will  usually  be  detected  in  milk  by  the 
dyeing  test  given  above,  though  complications  may 
sometimes  occur  owing  to  the  separation  of  the  colour 
with  the  curd  on  acidification.  The  following  method 
for  separating  the  colours  most  likely  to  occur  in  milk, 
and  tests  for  their  identification,  recommended  by 
Wynter  Blyth,  may  be  applied  with  advantage  :— 

If  the  fresh  milk  gives  a  pink  colour  with  hydrochloric 
acid,  the  presence  of  a  sulphonated  azo-dye  may  be 
inferred. 

If  a  piece  of  filter  paper  soaked  for  24  hours  in  the 
milk,  previously  make  alkaline  with  sodium  carbonate, 
takes  a  brown  stain  which  is  changed  to  pink  by  hydro- 
chloric acid,  the  presence  of  anatto  may  be  inferred. 

To  test  for  caramel,  10  c.c.  of  the  milk  are  coagulated 
by  acetic  acid,  and  the  curd  is  collected  by  straining 
through  linen,  transferred  to  a  white  porcelain  dish  and 
just  covered  with  concentrated  hydrochloric  acid  ;  a 
control  sample,  known  to  be  free  from  caramel,  is  treated 
in  the  same  manner.  A  blue  violet  coloration  on  adding 
the  acid  to  the  curd  in  the  dish  indicates  the  presence  of 
caramel. 

For  the  isolation  of  the  various  classes  of  colouring 
matters,  for  the  application  of  special  tests,  the  following 
method  is  recommended  :  At  least  60  c.c.  of  milk  are 
made  just  alkaline  to  delicate  litmus  paper  by  the 
addition  of  dilute  soda  or  strontia  solution,  and  evapo- 
rated to  a  thin  paste  on  the  water  bath.  The  paste  is 
thoroughly  extracted  with  ether,  which  will  remove  the 
fat  and,  if  the  milk  is  sour,  the  products  of  those  dyes 
which  have  been  reduced  by  nascent  hydrogen  produced 


Preservatives,  etc.,  in  Foods     315 

by  certain  bacteria  (compare  Reductase  Test,  p.  218). 
The  ether  is  evaporated  off,  and  the  fat  shaken  with  warm 
water.  The  water  is  separated  and  evaporated  to  dryness 
in  a  white  porcelain  dish  ;  a  coloured  residue  will  be  due 
to  added  colour,  which  may  possibly  be  identified  by  the 
special  tests  given  in  Wynter  Blyth's  "  Foods  :  Their 
Composition  and  Analysis,"  1909,  p.  242. 

The  fat-free  residue  is  extracted  with  boiling  alcohol, 
which  is  filtered  and  evaporated  to  dryness  in  a  white 
porcelain  dish.  If  the  residue  is  yellow  or  orange,  a 
portion  of  it  is  taken  up  with  a  little  decinormal  acid, 
and  shaken  with  ether.  In  the  presence  of  the  following 
dyes,  the  ether  will  be  coloured  :  natural  colours,  such 
as  anatto,  saffron,  turmeric,  etc.,  and  acid  coal-tar  dyes, 
including  the  acid  azo-dyes,  but  not  the  sulphonated 
azo-dyes.  The  latter,  as  well  as  the  basic  dyes,  do  not 
colour  the  ether. 

Coal-tar  dyes  may  be  recognised  by  the  dyeing  test 
described  above,  either  in  the  aqueous  layer,  or  in  an 
aqueous  solution  of  the  residue  obtained  by  evaporating 
the  ethereal  layer  to  dryness.  In  the  latter  it  will  not 
be  necessary  to  look  for  basic  dyes,  but  in  the  former 
either  acid  or  basic  dyes  may  be  found. 

Wynter  Blyth  recommends  the  following  tests  for 
anatto  in  the  residue  obtained  by  evaporating  the 
ethereal  layer,  as  supplementary  to  the  preliminary  test 
described  above  :  (i.)  A  drop  of  the  colouring  matter 
dissolved  in  water  and  made  alkaline  with  potash  gives 
an  orange  stain  on  filter  paper,  which  is  changed  to  pink 
by  stannous  chloride  solution,  (ii.)  A  little  of  the 
colouring  matter  is  dissolved  in  water  containing  a  little 
alcohol  and  a  drop  of  ammonia  ;  a  bundle  of  white  cotton 
fibres  is  introduced,  and  the  liquid  evaporated  nearly  to 
dryness.  The  fibre  is  then  immersed  in  a  solution  of 


316      Industrial  Organic  Analysis 

citric  acid  ;  it  will  be  coloured  rose  red  if  anatto  is 
present. 

Colouring  Matters  in  Meat. — Meat  is  generally  coloured 
with  coal-tar  dyes,  such  as  fuschine,  eosin  and  benzo- 
purpurin.  These  may  be  recognised  by  extracting  with 
alcohol,  and  applying  the  dyeing  test  to  the  extract  as 
described  above.  Carmine  and  cochineal  have  also  been 
used  for  colouring  meat. 

The  following  method,  due  to  Klinger  and  Bujard,  and 
modified  by  Bremer,  may  be  used  for  the  detection  of 
carmine  :  20  grams  of  the  minced  meat  are  heated  on 
the  water  bath  for  several  hours  with  a  mixture  of  equal 
parts  of  glycerol  and  water  which  has  been  slightly 
acidulated  with  tartaric  acid  ;  the  liquid  is  separated  by 
straining  through  muslin,  and  when  cold,  filtered  in 
order  to  separate  it  from  fat  and  suspended  matter.  On 
adding  alum  solution  and  then  ammonia,  the  precipitated 
aluminium  hydroxide  will  carry  the  colour  with  it,  as  a 
lake  ;  the  latter  is  filtered  off,  washed  with  water, 
dissolved  in  a  small  quantity  of  tartaric  acid  solution, 
and  examined  in  the  spectroscope  ;  if  carmine  is  present 
the  solution  will  show  absorption  bands  at  b  and  E,  and 
another  close  to  D,  in  the  solar  spectrum. 

Colouring  Matters  in  Wine. — Coal-tar  dyes  will  be 
detected  by  the  dyeing  test,  described  above  ;  in  addition 
to  these,  such  materials  as  elderberry,  logwood  and 
cochineal  are  said  to  be  in  use  for  colouring  wine.  The 
subject  is  fully  treated  of  in  Wynter  Blyth's  "  Foods : 
Their  Composition  and  Analysis,"  as  well  as  in  some  of 
the  other  works  mentioned  at  the  end  .of  this  chapter. 
The  tests  used  at  the  Municipal  Laboratory  of  Paris  for 
detecting  artificial  colouring  matters  in  wine  are  given 
on  p.  458  of  the  above-mentioned  work  (1909  edition). 

Detection  and  Estimation  of  Metallic  Colouring  Matters. 


Preservatives,  etc.,  in  Foods     317 

—The  commonest  metallic  colouring  matter  used  in 
foods  is,  perhaps,  copper  sulphate,  which  is  added  to 
preserved  peas,  spinach,  etc.,  to  impart  a  bright  green 
colour  and,  at  the  same  time,  to  harden  the  integument. 

A  general  method  for  the  detection  of  compounds  of 
copper,  arsenic,  lead,  chromium,  zinc,  etc.,  all  of  which 
must  be  regarded  as  poisonous,  is  as  follows  :  A  known 
weight,  say  100  grams  of  the  material,  in  the  case  of 
preserved  vegetables  or  fruits,  strained  before  weighing, 
is  burnt  off  at  a  low  red  heat,  the  residue  moistened  with 
concentrated  hydrochloric  acid,  and  heated  with  a  little 
water  on  a  boiling-water  bath  ;  after  filtering  and 
washing,  the  insoluble  residue  is  thoroughly  incinerated 
at  a  bright  red  heat,  dissolved  in  dilute  hydrochloric 
acid,  and  the  solution  added  to  the  nitrate  obtained  from 
the  previous  operation.  The  metals  may  then  be 
detected  and  estimated  in  the  solution  by  the  ordinary 
methods  of  inorganic  analysis.  The  heating  of  the 
material  should  be  carried  out  in  a  porcelain,  and  not  in 
a  platinum  dish,  as  the  latter  may  be  attacked  by  the 
foreign  metals  present. 

Arsenic,  if  present,  will  usually  have  been  introduced 
as  an  impurity  with  organic  or  mineral  colouring  matters, 
or  other  materials  used  in  the  manufacture  or  preparation 
of  the  food,  and  is,  therefore,  only  likely  to  be  present  in 
very  small  amounts  ;  and  must  accordingly  be  tested  for 
in  the  original  material  by  special  methods,  such  as  the 
Marsh  test ;  methods  for  the  detection  of  arsenic  in 
foods  will  be  found  fully  described  in  Wynter  Blyth's 
"  Foods  :  Their  Composition  and  Analysis,"  and  some  of 
the  other  works  mentioned  at  the  end  of  this  chapter. 

Ferric  oxide  is  said  to  be  added  to  cocoa,  in  order  to 
improve  its  colour,  and  as  an  adulterant ;  it  may  be 
detected  and  estimated  on  the  same  lines  as  described 
above  for  other  mineral  additions  to  foods. 


318      Industrial  Organic  Analysis 


REFERENCES. 

"  Preservatives  in  Food  and  Food  Analysis,"  by  Thresh  and 
Porter.  1906,  J.  &  A.  Churchill,  London. 

"  Foods:  Their  Composition  and  Analysis,"  by  A.  and  M. 
Wynter  Blyth.  1909,  Chas.  Griffin,  London. 

"  Foods  and  their  Adulteration,"  by  Harvey  Wiley.  1907, 
J.  &  A.  Churchill,  London. 

"  Select  Methods  in  Food  Analysis,"  by  Leffmann  and  Beam. 
1905,  Rebman,  Ltd.,  London. 

"  Food  Inspection  and  Analysis,"  by  Leach.  1909,  Wiley 
&  Sons,  New  York. 

"  The  Sale  of  Foods  and  Drugs  Act,"  by  J.  W.  Bell.  1910, 
Butterworth  &  Co.,  London. 

"  Foods  and  Drugs,"  by  E.  Parry.  1911,  Scott  &  Green- 
wood, London. 

"  The  Identification  of  Pure  Organic  Compounds,"  Vol.  III. 
(Colours),  by  S.  P.  Mulliken.  1910,  Wiley  &  Sons,  New  York. 

Allen's  "  Commercial  Organic  Analysis,"  Vol.  V.  (Colours). 
1910,  J.  &  A.  Churchill,  London. 

"  Chemisch-Technische  Analyse,"  Vol.  II.,  by  Post.  1908, 
Fr.  Vieweg,  Brunswick. 

Reports  to  the  Local  Government  Board  on  Public  Health 
and  Medical  Subjects,  New  Series,  No.  12,  Food  Report,  No.  12  : 
Dr.  G.  S.  Buchanan  and  Dr.  S.  B.  Schryver  on  the  "  Application 
of  Formaldehyde  to  Meat." 

Ditto,  New  Series,  No.  13,  Food  Report,  No.  10  :  Dr.  J.  M. 
Hamill  on  the  "  Use  of  Preservatives  in  Cream."  Wyman  & 
Sons,  London. 

Local  Government  Board  Reports  of  Inspector  of  Foods, 
No.  6:  "  On  Preservatives  in  Meat  Foods  packed  in  Cans  or 
Glass."  Wyman  &  Sons,  London. 

Statutory  Rules  and  Orders,  1912,  No.  1152 :  The  Public  Health 
(Milk  and  Cream)  Regulations. 


INDEX 

ACCURACY  of  results,  2 
Acid  in  burning  oil  or  kerosene,  169 
in  lubricating  oils,  101,  170 
value  of  fats,  80,  96 

estimation  of,  101 
Acidity  of  flour,  estimation  of,  260 
of  malt,  estimation  of,  278 
of  milk,  estimation  of,  194 
Adulteration  of  flour,  262 
of  foods,  282 

of  lubricating  oils,  170,  178 
of  milk,  208,  212 
Albumin  in  milk,  181,  186 

estimation  of,  203 
Alcohol  from  starch,  232,  244 

in  water,  beer,  spirits,  etc.,  estimation  of,  245,  247 
Alkali  in  soap,  127 

estimation  of,  133,  143 
Alkaline  carbonate,  silicate,  etc.,  in  soap,  estimation  of,  142 

hydroxide  in  soap,  estimation  of,  142 
Alum  in  flour,  260 

detection  of,  262,  263 
Alumina  cream,  226 
Amylan,  240 

Anatto  in  milk,  detection  of,  314,  315 
Anthracene,  estimation  of,  74 

oil  products,  carbazole  in,  estimation  of,  76 

composition  and  properties  of,  73 
paraffins  in,  estimation  of,  77 
Arachidic  acid,  79 

estimation  of,  112 
Arachis  oil,  Bellier's  test  for,  112 
estimation  of,  112 
in  cacao  butter,  116 
in  lard,  116 


320  Index 


Arachis  oil,  in  olive  oil,  in 
rape  oil  in,  116 

Aromatic  hydrocarbons  in  petroleum  naphtha  and  its  products, 

detection  of, 
163 

estimation  of,  164 
in  petroleum,  154,  156 
Arsenic,  detection  of,  in  foods,  317 
Ash  in  barley,  estimation  of,  259,  261 

in  burning  oil  or  kerosene,  estimation  of,  169 
in  butter,  189 

and  margarine,  estimation  of,  230 
in  coal,  estimation  of,  17 
in  flour,  255 

estimation  of,  259 

in  lubricating  oils,  estimation  of,  1 76 
in  malt,  estimation  of,  278 

extract,  estimation  of,  278 
Astatki,  155,  157 


BARLEY,  fat  in,  estimation  of,  271 
malt  from,  248 
mechanical  tests  for,  270 
nitrogen  in,  estimation  of,  271 
properties  and  composition  of,  248,  270 
proteins  in,  estimation  of,  271 
sampling  of,  270 

starch  in,  estimation  of,  236,  244 
sulphurous  acid  in,  detection  of,  272 
Beer,  alcohol  in,  estimation  of,  247 
benzoic  acid  in,  296 

detection  and  estimation  of,  304 
fluoride  in,  296 

detection  of,  302 
from  starch,  232 
saccharin  in,  296 

detection  and  estimation  of,  306 
salicylic  acid  in,  295 

detection  and  estimation  in,  304 
sulphites  in,  296 

detection  and  estimation  of,  300 
Benzene.     See  under  "  Benzols." 


Index  321 


Benzine  (see  also  under  "  Petroleum  naphtha  and  products  "), 

155,  157 
Benzole  acid  in  beer,  detection  and  estimation  of,  304 

in  beverages,  detection  and  estimation  of,  304 
in  butter  and  margarine,  detection  of,  290 
in  milk,  detection  of,  289 
in  wine,  detection  of,  304 
reactions  for,  289,  290,  306 
Benzols,  bromine  absorption  test  for,  52 

carbon  disulphide  in,  estimation  of,  48 

composition  and  properties  of,  45,  46 

distinction  from  petroleum   spirit  and  shale  naphtha, 

55 

fractional  distillation  test  for,  47 
non-nitratable  hydrocarbons  in,  estimation  of,  50 
paraffins  in,  46 
specific  gravity  of,  47 
sulphuric  acid  test  for,  53 
Bicarbonate  of  soda  in  flour,  260,  264 

in  milk,  285 
Boric  acid  or  borax  in  butter  and  margarine,  189,  285 

detection  and  estimation  of,  230,  287 
in  cream,  283 

detection  and  estimation  of,  286,  287 
in  fruit  juices,  296 

detection  and  estimation  of,  299 
in  jams,  296 

detection  and  estimation  of,  299 
in  meat,  etc.,  294 
detection  of,  297 
estimation  of,  299 
in  milk,  detection  of,  286 

estimation  of,  287 
in  sausages,  295 
in  wine,  296,  299 
reactions  for,  286,  298 
Bran,  in  flour,  253,  255,  259 
Burning  oil,  154,  157 

acid  in,  estimation  of,  169 

ash  in,  estimation  of,  169 

degree  of  refinement  of,  estimation  of,  167 

distillation  test  for,  166 

flash  point  of,  167 


322  Index 

Burning  oil,  impurities  in,  estimation  of,  168 

specific  gravity  of,  166 

sulphur  in,  estimation  of,  168 
Butter.     See  also  under  "  Butter  and  margarine." 

ash  in,  189  ' 

composition  of,  184,  188, 

fat,  119,  184,  188 

from  cream,  184,  187 

lecithin  in,  188 

margarine  in,  detection  and  estimation  of,  119 

pasteurising  test  for,  218,  231 

proteins  in,  189 

Sale  of,  Regulations  of  1902,  85 

salt  in,  189 

sampling  of,  227 
Butter  and  margarine,  ash  in,  estimation  of,  230 

benzoic  acid  in,  detection  of,  289 
boric  acid  in,  285 

estimation  of,  230,  288 
casein  in,  estimation  of,  229 
curd  in,  estimation  of,  229 
fat  in,  estimation  of,  228 

examination  of,  119,  230 
fluoride  in,  detection  of,  293 

in  presence  of  boric  acid,  293 
non-fatty  solids  in,  estimation  of,  229 
proteins  in,  estimation  of,  229 
salicylic  acid  in,  estimation  of,  290 
salt  in,  estimation  of,  230,  288 
sampling  of,  227 

CACAO  butter,  arachis,  sesame  or  cotton  oil  in,  117 
cocoanut  or  palm  kernel  oil  in,  118 
paraffin  wax  in,  122 

Calcium  acid  phosphate  in  flour,  262,  264 

detection  of,  264 
sulphate  in  flour,  264 

Calorific  value  of  coal,  29 

Cane  Sugar.     See  under  "  Sucrose." 

Caramel  in  milk,  detection  of,  314 

Carbazole  in  anthracene  oil  products,  estimation  of,  76 

Carbolic  acid,  crude,  composition  of,  57,  64 
liquefied,  composition  of,  60 


Index  323 


Carbolic  acid,  liquefied,  phenol  in,  estimation  of,  61 
liquid,  composition  of,  61 

phenol  in,  estimation  of,  58 
tests  for,  6 1 
pure  crystallised,  composition  of,  60 

phenol  in,  estimation  of,  61 
solidifying  point  of,  63 
tests  for,  6 1 

water  in,  estimation  of,  64 
specific  gravity  of,  36,  56 
total  phenols  and  water  in,  estimation  of,  57 
oil,  composition  of,  56 

naphthalene  in,  estimation  of,  57 
specific  gravity  of,  36,  56 
soaps,  phenols  and  neutral  hydrocarbons  in,  estimation 

of,  146 
Carbon  disulphide  in  benzols,  46 

estimation  of,  48 
free,  in  coal  tar,  estimation  of,  37 
in  coal,  8 

estimation  of,  27 
Carbonate  alkaline  in  milk,  285 
in  soap,  142 
Casein,  composition  and  properties  of,  185 

in  butter  and  margarine,  estimation  of,  229 
in  milk,  estimation  of,  202 
Catalase,  217 

Caustic  alkali  (free)  in  soaps,  estimation  of,  142 
Chloride  (sodium)  in  butter  and  margarine,  estimation  of,  230, 

288 

in  meat,  estimation  of,  303 
Chlorides  in  soap,  estimation  of,  134 
Coagulation  of  milk,  186,  187 
Coal,  ash  in,  estimation  of,  17 
calorific  value  of,  29 
carbon  in,  8 

estimation  of,  27 
coke  from,  18 

combustible  residue  from,  estimation  of,  18 
composition  and  classification  of,  7 
distillation  products  of,  33 
hydrogen  in,  8 

estimation  of,  27 

21—2 


324  Index 

Coal,  nitrogen  in,  estimation  of,  21 
oxygen  in,  28 

physical  examination  of,  31 
sampling  of,  12 
sulphur  in,  estimation  of,  20 

volatile  combustible  matter  from,  estimation  of,  18 
water  in,  estimation  of,  18 
Coal  tar,  colours  in  foods,  310 

detection  of,  311 
in  milk,  313 

detection  of,  314 
composition  and  properties  of,  33 
distillation,  products  of,  35 

test  for,  38 

free  carbon  in,  estimation  of,  37 
oils  in  lubricating  oils,  170 
specific  gravity  of,  36 

or  mineral  oils,  detection  of,  180 
Cocoa,  ferric  oxide  in,  317 
Cocoanut  oil  in  cacao  butter  and  lard,  118 

•  "  Reichert  value  of,  100 

Coke,  ash  in,  estimation  of,  17 
from  coal,  18 

physical  examination  of,  31 
sampling  of,  12 
sulphur  in,  estimation  of,  20 
volatile  combustible  matter  from,  20 
water  in,  estimation  of,  16 
Cold  test  for  lubricating  oils,  174 
Combustible  residue  from  coal,  estimation  of,  18 
Colours  (artificial)  in  foods,  310 

detection  of,  311 
in  meat,  detection  of,  316 
in  milk,  313 

detection  of,  314 
in  wine,  detection  of,  316 
metallic,  in  foods,  detection  of,  317 
Condensed  milk,  cane  sugar  in.     See  "  Sucrose." 
carbohydrates  in,  222 
composition  of,  221 
concentration  of,  226 
invert  sugar  in,  222 
estimation  of,  226 


Index  325 


Condensed  milk,  lactose  in,  222,  225 
sucrose  in,  221 

estimation  of,  223 
test  for,  223 

total  sugar  in,  calculation  of,  223 
Copper  sulphate  in  flour,  260 

detection  of,  263 
in  foods,  310,  311 

detection  of,  317 
Cotton  oil,  detection  of,  109 

in  cacao  butter,  117 
in  lard,  117 
in  olive  oil,  109 
in  rape  oil,  116 
Cracked  oils,  156 

Cream,  boric  acid  in,  detection  and  estimation  of,  286 
butter  from,  184,  187 
fat  in,  estimation  of,  197,  198,  200 
t    fluorides  in,  detection  of,  293 
from  milk,  184 
pasteurising  of,  187 

Storch's  test  for,  217 
preservatives  in,  283 
removal  of,  from  milk,  210,  211,  213 
souring  of,  184,  187 
Creoline,  analysis  of,  150 

composition  of,  149 

Creosote  oil,  composition  and  properties  of,  65 
Creosoting  liquor,  basic  constituents  in,  estimation  of,  68 
description  of,  68 
phenols  in,  estimation  of,  69 
Cresols,  56,  60,  63 

estimation  of,  69 
Crude  fibre  in  flour,  255 

estimation  of,  261 
Curd  in  butter,  189 

and  margarine,  estimation  of,  229 
Curdling  of  milk,  186,  187 
Cylinder  oils,  157 


DEPARTMENTAL    Committee    on    Preservatives    and    Colouring 
Matters  in  Foods,  Report,  1901.  .295,  311 


326 


Index 


Dextrin,  estimation  of,  242 

from  starch,  237 

Diastase,  action  of,  on  starch,  232,  235,  237,  238 
preparation  of,  238 
and  yeast,  action  of,  on  starch,  244 
Dirt  in  milk,  estimation  of,  188,  192 
Distillation  test  for  benzols,  47 

for  burning  oil  or  kerosene,  166 
for  coal  tar,  38 

for  first  runnings  and  light  oils,  43 
for  petroleum,  158 

naphtha  and  products,  163 
Dyeing  method  for  detecting  coal  tar  colours,  311 


ENGLER  distillation  test,  160 

Enzymes  in  milk,  188,  217,  218 

Ergot,  flours  damaged  by,  265,  266 

Ester  value  of  fats,  96 

Extract,  estimation  of,  in  malt,  275 
from  malt  for  analysis,  274 

for  Lintner  value,  250 
for  starch  conversion,  238 


FAT  in  barley,  estimation  of,  271 
in  butter,  184 

and  margarine,  estimation  of,  228 

examination  of,  119 
in  cream,  estimation  of,  197,  198,  200 
in  flour,  255 

estimation  of,  260 
in  milk,  183,  211 

estimation  of,  195,  198 
(unsaponified)  in  soap,  estimation  of,  130 
Fatty  acids  from  fats,  78,  80 

and  rosin  acids  in  soap,  estimation  of,  131,  145 

in  fats,  estimation  of,  101 

in  sheep  dips  and  phenolic  disinfectants,  estimation 

of,  150 

in  soap,  estimation  of,  142 
solidifying  point  of,  89 


Index  327 

Fatty  oils  and  fats,  acid  value  of,  80,  96,  101 
classification  of,  81 
composition  and  properties  of,  79 
constants,  of  106 
ester  value  of,  96 
estimation  of,  82 
examples  of  analysis  of,  105 
iodine  value  of,  91 
melting  point  of,  87 
physical  examination  of,  86 
phytosteryl  acetate  test  for,  104 
preparation  of,  for  analysis,  84 
Reichert-Wollny  value  of,  98 
saponification  of,  79,  89 

value  of,  95 
specific  gravity  of,  87 
titer  test  for,  89 
unsaponifiable    matter    in,    isolation    of, 

102 

vegetable,  detection  of,  104 
water  in,  estimation  of,  85 
in  lubricating  oils  and  mineral  oils,  170,  177 
Fehling's  solution,  139,  206 
Fermenting  test  for  milk,  218 
First  runnings,  distillation,  products  of,  45 

test  for,  43 

phenols  in,  estimation  of,  44 
properties  and  composition  of,  40 
specific  gravity  of,  41 
Fish  and  blubber  oils  in  rape  oil,  118 
Fixed  carbon  in  coal,  18 

combustible  residue  from  coal,  estimation  of,  18 
Flash  point  of  burning  oil  or  kerosene,  167 
Fleischmann's  formula,  210 
Flours,  foreign,  detection  of,  in  wheat  flour,  266 
germ,  253,  255,  258 
self-raising,  260,  264 
special,  253 

starch  in,  estimation  of,  236,  244 
wheat,  acidity  of,  260 

alum  in,  detection  of,  260,  262,  263,  265 
ash  in,  255 

estimation  of,  259 


328 


Index 


Flours,  wheat,  bicarbonate  of  soda  added  to,  260,  264 
classification  of,  252 
composition  and  properties  of,  255 
copper  sulphate  in,  260,  263 
crude  fibre  in,  255 

estimation  of,  261 
damaged  by  ergot,  detection  of,  265,  266 

moulds,  detection  of,  265,  266 
fat  in,  255 

estimation  of,  260 
foreign  flours  in,  257,  262 
detection  of,  262,  266 
gliadin  in,  256 

estimation  of,  258 
gluten  in,  256 

test  for,  256 
glutenin  in,  256 

grades  and  qualities,  various,  252,  253,  259 
improvers  in,  260,  264 
leguminous  starches  in,  detection  of,  268 
mineral  matter  added  to,  260,  262 

detection  of,  262 
nitrogenous  matter  in,  255 
estimation  of,  257 
soluble,  estimation  of,  258 
proteins,  total,  estimation  of,  257 
Fluorides,  detection  of,  293 

in  presence  of  boric  acid,  293 
in  beer,  meat  and  wine,  detection  of,  302 
Foods,  colouring  matters  used  in,  310 
preservatives  used  in,  281 
preserving  of,  280 

See  also  under  "  Milk,"  "  Meat,"  "  Wine,"  etc. 
Foreign  bodies  in  milk,  188 

flours  in  wheat  flour,  257,  262,  266 
Formaldehyde  in  meat,  294 

detection  and  estimation  of,  296 
in  milk,  detection  of,  291 
Formic  acid  in  fruit  juices,  jams,  etc.,  296 
detection  of,  307 
estimation  of,  308 

Fruit  juices  and  jams,  formic  acid  in,  detection  and  estimation 
of,  307,  308 


Index  329 


Fruit  juices  and  jams,  sulphites  in,  detection  and  estimation  of, 
300 

See  also  under  "  Meat  "  and  "  Wine." 


GASOLENE  (see  also  under  "  Petroleum  naphtha  and  products"), 

155 
Gliadin  in  flour,  256 

estimation  of,  258 
Globulin  in  flour,  256 
in  milk,  185 

Glucose,  estimation  of,  137,  141 
Gluten  in  flour,  256 
Glycerol  in  soap,  estimation  of,  134 

in  presence  of  sugar,  136 
Gumming  of  lubricating  oils,  170,  180 


HYDROCARBONS  in  carbolic  soaps,  estimation  of,  146 

in  sheep  dips  and  phenolic  disinfectants,  estima- 
tion of,  150 
Hydrogen  in  coal,  estimation  of,  27 

peroxide  in  milk,  detection  of,  292 


IMPROVERS  in  flour,  260,  264 
Insoluble  matter  in  soap,  127,  142,  143 
Invert  sugar,  estimation  of,  137,  141,  226 
Iodine  value  of  fats,  91 


JAMS,  preservatives  in  (see  also  under  "  Meat  "  and  "  Wine  "), 

296 
formic  acid  in,  296 

detection  and  estimation  of,  307,  308 


KEROSENE  (see  also  under  "  Burning  oils  "),  155,  157,  165 
Kjeldahl  method  for  determining  nitrogen,  22 

Gunning's  modification,  26 
Kottstorfer  value,  102 


33°  Index 

LACTIC  acid  in  butter,  189 

in  milk,  182,  186,  194 
Lactometer,  191 
Lactose  in  butter,  189 

in  condensed  milk,  222,  225 
in  milk,  186 

estimation  of,  204,  206 
fermentation  of,  186 

Lard,  arachis,  sesame  and  cotton  oils  in,  117 
cocoanut  or  palm  kernel  oil  in,  118 
paraffin  wax  in,  128 
Lard  oil  in  olive  oil,  1 1 1 
Lecithin  in  milk  and  butter,  188 
Legumin,  separation  of,  268 

Leguminous  starches,  detection  of,  in  wheat  flour,  268 
Light  oils  (of  tar),  composition  and  properties  of,  40 
distillation,  products  of,  45 

test  for,  43 

phenols  in,  estimation  of,  44 
0  specific  gravity  of,  41 

See  also  under  "  Petroleum  naphtha  and 

products." 

of  petroleum,  155,  157 

Lignite  oils  in  lubricating  oils  and  mineral  oils,  170,  181 
Lintner  value  of  malt,  249 
Local  Government  Board  Circular  Letter,  July,  1906.  .284 

Reports.     See  references  after  Chap- 
ters VII.  and  VIII. 
Lubricating  greases,  lignite  oils  in,  181 

soap  in,  estimation  of,  176 
soda  or  other  bases  in,  estimation  of,  1 76 
tar  oils  in,  180 
Lubricating  oils,  acid  in,  102,  ijo 

estimation  of,  175 
adulteration  of,  170,178 
ash  in,  176 
classification  of,  171 
cold  test  for,  174 
cylinder,  157 
fatty  oils  in,   170 

estimation  of,  177 
flash-point  of,  172 
from  petroleum,  154,  156,  157 


Index  331 


Lubricating  oils,  gumming  of,  170,180 

rosin  or  rosin  oils  in,  170,175 

detection  of,  178,  179 
setting  point  of,  173 
soap  and  ash  in,  estimation  of,  176 
specific  gravity  of,  174 
viscosity  of,  170 
water  in,  170 

estimation  of,  175 

Lysol,  composition  and  analysis  of,  149,  150 


MALT,  acidity  of,  278 

ash  in,  estimation  of,  278 

diastatic  value  of,  249 

extract  from,  estimation  of,  275 

preparation  of  for  analysis,  274 
maltose  in,  estimation  of,  277 
mechanical  tests  for,  273 
nitrogen  in,  estimation  of,  277 
production  of,  from  barley,  248 
proteins  in,  estimation  of,  277 
saccharification  test  for,  274 
sampling  of,  273 

starch  in,  estimation  of,  236,  244 
sugar  in,  estimation  of,  277 
water  in,  estimation  of,  273 
wort,  preparation  of,  from,  275 

specific  gravity  of,  to  estimate  extract,  275 
Malt  extract,  ash  in,  estimation  of,  278 

maltose  in,  estimation  of,  277 
nitrogen  in,  estimation  of,  277 
preparation  of  for  analysis,  274 

Lintner  value,  250 
starch  conversion,  238 
proteins  in,  estimation  of,  277 
sugar  in,  estimation  of,  277 
Maltose,  estimation  of,  242 

in  malt,  estimation  of,  277 

extract,  estimation  of,  277 
Margarine,  in  butter,  detection  of,  119 
water  in,  estimation  of,  85 

See  also  "  Butter  and  margarine." 


332  Index 


Mechanical  tests  for  barley,  270 
for  coke,  31 
for  malt,  273 
Melting  point  of  fats,  87 

Metallic  colouring  matters  in  foods,  detection  of,  317 
Micro-organisms  in  milk,  186,  188,  190,  214,  218 
Middle  oil.     See  "  Carbolic  oil." 
Milk,  acidity  of,  182,  186,  194 

estimation  of,  194 
addition  of  water  to,  detection  of,  209,  212 

or  skim  milk  and  removal  of  cream , 

detection  of,  211,  213 
adulteration  of,  208,  212 
albumin  in,  185,  186 

estimation  of,  203 
anatto  in,  detection  of,  314,  315 
ash  in,  187,  285 

estimation  of,  208 
caramel  in,  detection  of,  314 
casein,  composition  and  properties  of,  185 

estimation  of,  202 

coal  tar  colours  in,  detection  of,  314,  315 
colouring  matters,  artificial,  in,  313 

detection  of,  314 

composition  and  properties  of,  182,  190 
condensed.     See  under  "  Condensed  milk." 
curdling  of,  186 

decomposition  of,  by  micro-organisms,  186,  214,  218 
dirt  in,  188 

estimation  of,  192 
enzymes  in,  188,  217,  218 
fat,  183,  211,  212 

estimation  of,  195,  198,  200 
fermentation  of,  186,  214,  218 
fermenting,  test  for,  219 
fluoride  in,  detection  of,  293 
foreign  bodies  in,  188 
formaldehyde  in,  detection  of,  291 
freshness  of,  194,  218 
globulin  in,  185 

hydrogen  peroxide  in,  detection  of,  292 
lactic  acid  in,  182,  186 
estimation  of,  194 


Index  333 


Milk,  lactose  in,  186 

estimation  of,  204,  206 

'';in  condensed  milk,  222,  225 
lecithin  in,  188 

micro-organisms  in,  186,  188,  214,  218 
mineral  matter  in,  187 
estimation  of,  208 
pasteurising  of,  214 

Storch's  test  for,  217 
preservation  of,  214 
preservatives  in,  283 
proteins  in,  185 

estimation  of,  201 

reductase  and  fermenting  test  for,  218 
removal  of  cream  from,  detection  of,  210,  213 

and    addition    of   water    or    skim 

milk,  detection  of,  211,  213 
saffron  in,  detection  of,  315 
Sale  of,  Regulations,  1901 . .  185 
(and  Cream),  Sale  of,  Regulations,  1912.  .283 
sampling  of,  189 
separated.     See  "  Skim  milk." 
separating  of,  184 
skim,  184. 

See  also  "  Skim  milk." 

sodium  carbonate  or  bicarbonate  in,  detection  of,  285 
solids,  187 

less  fat,  185,  212 

estimation  of,  207,  209 
solids,  total.     See  "  Milk  solids,  less  fat." 
souring  of,  182,  186 
specific  gravity  of,  191,  210 
sterilising  of,  214 
sugar.     See  "  Lactose." 

Mineral  matter  added  to  flour,  260,  262,  264, 
of  milk,  187 

See  also  "  ash  "  under  "  Milk." 
oils,  fatty  oils  in,  detection  and  estimation  of,  177 
lignite  oils  in,  181 
rosin  in,  178 

oil  in,  179 
tar  oils  in,  180 
Molisch  reaction  for  carbohydrates,  136,  185 


334  Index 


NAPHTHA,  coal  tar,  40,  45. 

See  also  under  "  Benzols  "  and  "  Light  oils." 
petroleum,  161. 

See     also     under     "  Petroleum      naphtha     and 

products." 
shale,  47 

distinction  from  benzol,  55 
Naphthalene,  tests  for  purity,  72 
Naphthenes  in  petroleum,  154,  156 
Naphthol,  detection  of,  310 
Nitrates  in  meat,  estimation  of,  303 
Nitrogen,  estimation  of,  by  Kjeldahl  process,  22 
in  coal,  21. 

See  also  under  "  Proteins." 
Nitrogenous  matter.     See  under  "  Proteins." 
Non-fatty  solids  in  butter  and  margarine,  estimation  of,  229 
Non-nitratable  matter  in  benzols,  estimation  of,  50 

OFFAL  of  flour,  253,  239 
Oil  cake,  83 

fat  in,  estimation  of,  83 
crude  fibre  in,  estimation  of,  261 
Olive  oil,  arachis  oil  in,  1 1 1 

cotton  oil  in,  109 

lard  oil  in,  1 1 1 

rape  oil  in,  115 

sesame  oil  in,  no 

Original  gravity  of  fermented  liquor,  determination  of,  247 
Oxygen  in  coal,  estimation  of,  28 

PALM  kernel  oil  in  cacao  butter  and  lard,  118 

Reichert  value  of,  100 
Paraffin  wax  in  cacao  butter  and  lard,  122 
Paraffins,  distinction  from  benzols,  55 
in  anthracene,  74 

estimation  of,  77 
in  benzols,  46 
in  petroleum,  153,  154,  156 
Pasteurising  of  cream,  187 
milk,  214 

and  cream,  Storch's  test  for,  217 
Pentosans  in  starch,  236 


Index  335 


Petrol,  155,  157. 

See  also  under  "  Petroleum  naphtha  and  products." 
Petroleum,  composition  and  properties  of  153 
crude,  distillation  test  for,  159 

sediment,  in  estimation  of,  158 
specific  gravity  of,  158 
water  in,  estimation  of,  158 
distillation  of,  154 
ether,  156. 

See     also     under     "  Petroleum     naphtha    and 

products." 

from  various  localities,  153,  154,  156 
naphtha  and  products,  161 

aromatic  hydrocarbons  in  estimation  of,  164 
distillation  test  for,  163 
flash-point  of,  163 
rosin  oils  in,  162,  164  . 
turpentine  in,  162,  164 
Phenol,  estimation  of,  56,  61,  63 
properties  of,  60 
solidifying  point  of,  63 
water  in,  estimation  of,  64 
Phenolic  disinfectants  containing  soap,  149 

analysis  of,  150 

powders,  analysis  of,  149 

Phenols  in  carbolic  soaps,  estimation  of,  146 

in  disinfecting  powders,  estimation  of,  65 
in  naphthalene,  detection  of,  72 
in  phenolic  disinfectants,  estimation  of,  150 
in  sheep  dips,  estimation  of,  150 
Phosphates  added  to  flour,  260,  264 

detection  of,  264 

Physical  examination  of  coal  and  coke,  31 
Phytosteryl  acetate  test  for  vegetable  fats,  104 
Potato,  composition  of,  255 

starch,  detection  of,  in  flour,  267 
starch  in,  estimation  of,  236,  244 
Preservation  of  foods,  280,  281,  294 
of  milk,  214 

of  samples  for  analysis,  190,  192 
Preservatives  in  butter  and  margarine,  189 
in  foods,  280,  281,  294 
in  meat,  wine,  etc.,  294 


336 


Index 


Preservatives  in  milk,  283 

and    colouring    matters    in    ioods,    Departmental 

Committee's  Report  on,  1901 .  .295,  311 
Proteins  in  barley,  estimation  of,  271 
in  butter,  189 

and  margarine,  estimation  of,  229 
in  flour,  estimation  of,  257 
in  malt,  estimation  of,  277 
in  milk,  185 

estimation  of,  201 
soluble  in  flour,  255,  256 

estimation  of,  258 
total,  in  flour,  255 

Public  Health  (Milk  and  Cream)  Regulation,  1912.  .283 
Pyridine  bases  in  creosoting.  liquor,  estimation  of,  70 

in  phenolic  disinfectants,  estimation  of,  150 
in  sheep  dips,  estimation  of,  150 

QUINOLINE  bases  in  creosoting  liquor,  estimation  of,  70 
in  disinfectants,  estimation  of,  150 
in  naphthalene,  detection  of,  73 
in  phenolic  disinfectants,  estimation  of,  150 
in  sheep  dips,  estimation  of,  150 

RAPE  oil  in  arachis  and  olive  oils,  115 
cotton  oil  in,  116 
fish  or  blubber  oils  in,  118 
Reductase  and  fermenting  test  for  milk,  218 
Reichert-Wollny  value  of  fats,  98 
Reporting  of  results,  4 
Rice  flour  in  wheat  flour,  269 
Rosin,  Liebermann  Storch,  reaction  for,  144 
Rosin  acids  in  soap,  126 

detection  and  estimation  of,  144 
in  phenolic  disinfectants  and  sheep  dips,  150 
or  rosin  oil  in  lubricating  oils,  170,  175,  178,  179 

in  petroleum  naphtha  products,  162,  164 

SACCHARIFICATION  test  for  malt,  274 
Saccharin  in  beverages,  296 

detection  and  estimation  of,  306 
separation  from  benzoic  and  salicylic  acids,  306 


Index  337 


Saffron  in  milk,  detection  of,  315 
Sale  of  Butter  Regulations,  1902.  .85 

Foods  and  Drugs   Act,    1875.. 210,  264,  266,   282,   284, 

285,  295,  3ii 

Milk  Regulations,  1901 . .  185 

Salicylic  acid  in  beverages,  detection  and  estimation  of,  304 
in  butter  and  margarine,  detection  of,  290 
in  milk,  detection  of,  289 
in  wines,  beers,  fruit  juices,  etc.,  295 
reactions  for,  289,  290,  306 
separation  of,  from   benzoic  and    salicylic    acids, 

306 

Salol,  detection  of,  310 
Salt,  in  butter,  189 

and  margarine,  estimation  of,  230,  288 
in  meat,  estimation  of,  303 
Sampling,  i 

of  barley,  270 
of  butter,  227 
of  coal,  12 
of  fats,  84 
of  malt,  273 
of  milk,  189 

Sanatol,  saprol,  solutol  and  salveol,  150 
Saponification  of  fats,  79,  89 

value,  95 

Sediment  in  petroleum,  158 
Separated  milk.     See  under  "  Skim  milk." 
Sesame  oil,  detection  of,  no 

in  cacao  butter  and  lard,  117 
olive  oil,  no 

Setting  point  of  lubricating  oils,  173 
Sheep  dips,  analysis  of,  71,  150 
Silicate  (alkaline)  in  soap,  estimation  of,  142 
Skim  milk,  addition  of,  to  milk,  211 
addition  of  water  to,  212 
fat  in,  184 

estimation  of,  196,  198,  200 
solids  less  fat  in,  212 
specific  gravity  of,  192 

Soaps,  alkaline  carbonate  and  silicate  in,  estimation  of,  142 
carbolic,  analysis  of,  146 
caustic  alkali  (free)  in,  estimation  of,  142 

l.O.A.  22 


338 


Index 


Soaps,  classification  of,  126 

chlorides  and  sulphates  in,  estimation  of,  134 

composition  and  properties  of,  125 

fat,  unsaponified,  in,  estimation  of,  130 

fatty  or  fatty  and  rosin  acids  in,  estimation  of,  131,  133, 

J45 
glycerol  in,  estimation  of,  134 

and  sugar  in,  estimation  of,  136 
in  creoline,  estimation  of,  151 
in  lubricating  greases,  176 
in  lubricating  oils,  176 
in  lysol,  estimation  of,  151 
in  phenolic  disinfectants  and  sheep  dips,  149 

estimation  of,  151 
in  sheep  dips,  estimation  of,  151 
insoluble  matter  in,  estimation  of,  127,  143 
results  of  analysis  of,  146 
rosin  acids  in,  126 

detection  and  estimation  of,  144 
scheme  of  analysis  of,  128 

soda,  combined  and  total  in,  estimation  of,  133,  143 
soluble,  fatty  acids  in,  estimation  of,  133 
sugar  in,  estimation  of,  137 
total  alkali  in,  estimation  of,  133,  143 
typical  analyses  of,  147 
water  in,  estimation  of,  130 
Soda  (or  other  bases)  in  lubricating  greases,  estimation  of,  176 

in  soap,  estimation  of,  133,  142,  143 
Sodium  carbonate  in  soap,  estimation  of,  142 

or  bicarbonate  in  milk,  detection  of,  285 
Solidifying  point  of  fatty  acids,  90 

of  lubricating  oils,  173 
of  phenol,  63 

Solids  less  fat  in  butter  and  margarine,  estimation  of,  229 
in  milk,  209,  211 

estimation  of,  210 
Souring  of  milk,  182,  186,  217 
Soxhlet's  extraction  apparatus,  66 

lactometer,  191 

Specific  gravity  of  alcohol  and  water  mixtures,  246 
of  burning  oil  or  kerosene,  166 
of  carbolic  or  middle  oil,  56,  36 
of  coal  tar,  36 


Index  339 

Specific  gravity  of  fats,  87 

of    first    runnings,   light    oils    and    total    light 

oils,  41 

of  lubricating  oils,  174 
of  malt  extract,  276 
of  milk,  191,  210 

of  petroleum  naphtha  and  products,  161 
Westphal  balance  for  determining,  42 
Spirits,  alcohol  in,  estimation  of,  247 

from  starch,  232,  244 
Starch,  alcohol  and  spirits  from,  232 
beer  from,  232 
decomposition  of  by  diastase  and  yeast,  244 

estimation  of,  235,  236,  244 
hydrolysis  of,  by  acids,  232,  235 

diastase,  232,  235,  237 
in  barley,  estimation  of,  236,  244 
in  flour,  estimation  of,  236,  244 
occurrence  and  uses  of,  232 
pentosans  accompanying,  236 
properties  and  identification  of,  233 
Sterilising  of  milk,  214 
Streptococcus  lacticus,  196 
Sucrose,  estimation  of,  137 

in  condensed  milk,  221 

estimation  of,  222,  223 
in  soap,  127 

estimation  of,  137 
test  for,  in  presence  of  lactose,  223 
Sugars,  cane  and  invert,  estimation  of,  137,  141 
Sulphates  in  soap,  estimation  of,  134 
Sulphites  in  foods  and  beverages,  294 
detection  and  estimation  of,  300 
Sulphur  in  burning  oils,  157,  168 

in  coal  and  coke,  estimation  of,  20 

dioxide,  bleaching  of  barley  by,  test  for,  272 


TAR  oils  in  lubricating  oils,  181 

Titer  test,  89 

Total  alkali  in  soap,  estimation  of,  133,  143 

Turmeric  in  milk,  detection  of,  315 

Turpentine  in  petroleum  naphtha  products,  162,  164 


340  Index 


UNITED  States   of  America,   laws  regarding  preservatives  and 

colouring  matters  in  foods,  283,  285,  294,  311 
Unsaponifiable  matter  in  fats,  102 

VOLATILE  combustible  matter  from  coal,  estimation  of,  18 

coke,  20 
fatty  acids,  estimation  of,  98 

WATER  added  to  milk,  detection  of,  209,  212 
skim  milk,  detection  of,  212 

in  barley,  estimation  of,  271 

in  butter  and  margarine,  estimation  of,  85 

in  carbolic  acid,  estimation  of,  57 

in  coal,  estimation  of,  13 

in  coke,  estimation  of,  16 

in  fats,  estimation  of,  85 

in  flour,  estimation  of,  259 

in  lubricating  oils,  170,  175 

in  malt,  estimation  of,  271,  273 

in  petroleum,  estimation  of,  158 

in  phenol,  estimation  of,  64 

in  soaps,  estimation  of,  130 
Westphal  balance,  42 
Wheat  flour.     See  "  Flours,  wheat." 
Wine,  colouring  matters  in,  detection  of,  316 
preservatives,  various,  in,  294,  295,  296 

detection  and  estimation  of,  300,  302,  304,  306 
Wort,  preparation  of,  from  malt,  274 

specific  gravity  of,  for  extract  in,  275 

XYLENE,  composition  of,  53 
in  benzols,  45,  46 
meta,  estimation  of,  54 
paraffins  in,  estimation  of,  54 

YEAST,  action  of,  on  cane-sugar,  225 
on  dextrin,  244 
on  maltose,  233,  244 


BRADBURY,    AGNEW,   &  CO.    LD.,    PRINTERS,    LONDON   AND   TONBRIDGE. 


l-v 

ft 


fa 

i 
i 


OS     4 


University  of  Toronto 
Library 


DO  NOT 

REMOVE 

THE 

CARD 

FROM 

THIS 

POCKET 


Acme  Library  Card  Pocket 

Under  Pat.  "Ref.  Index  File" 
Made  by  LIBRARY  BUREAU